Content of Theory, Research and Development in our journal

Published in last 1 year |  In last 2 years |  In last 3 years |  All Please wait a minute... For Selected: Download Citations EndNote Ris BibTeX Toggle Thumbnails Select ReliefF Weighted Neighborhood Rough Sets and Attribute Reduction Based on Random Multi-Attribute Subspaces WANG Li Computer Engineering and Applications    2024, 60 (8): 69-77.   DOI: 10.3778/j.issn.1002-8331.2307-0232 Abstract （14）      PDF（pc） （572KB）（14）       Knowledge map    Save Attribute reduction is an important preprocessing method for data dimensionality reduction, but most existing attribute reduction methods do not consider the information of attribute weights in information systems. The ReliefF algorithm is a simple and efficient method for evaluating attribute weights. A ReliefF weighted neighborhood rough set and attribute reduction algorithm based on random multi-attribute subspace is proposed in this paper. Firstly, this method generates multiple sets of attribute set partitions with the same size random subspaces. The local weights of attributes in each set of partitioned random subspaces are calculated using the ReliefF algorithm, and the average of the local weights of attributes obtained from all sets is calculated to obtain the final global weights of each attribute in the information system. Then, based on the results of attribute weights, a new weighted neighborhood rough set model is proposed, and the related theories and properties are proved. Finally, based on this model, an attribute reduction algorithm for information systems is proposed by weighting neighborhood dependency. The experimental results of attribute reduction on public datasets show that the proposed algorithm has better reduction performance than the existing algorithms of the same type. Reference | Related Articles | Metrics Select Hybrid Driven Particle Swarm Algorithm CHEN Feng, DING Quan, WU Le, LIU Aiping, CHEN Xun, ZHANG Yunfei Computer Engineering and Applications    2024, 60 (8): 78-89.   DOI: 10.3778/j.issn.1002-8331.2307-0368 Abstract （28）      PDF（pc） （656KB）（26）       Knowledge map    Save The particle swarm optimization(PSO) algorithm is a widely applied optimization algorithm in fields such as robot motion planning and signal processing. However, this algorithm is prone to getting stuck in local optima, resulting in premature convergence problem. One reason for this premature convergence problem is that the particle swarm relies solely on fitness values to select learning examples. To overcome this problem, a particle swarm optimization algorithm called FINPSO(particle swarm optimization algorithm based on a hybrid approach driven by fitness values, improvement rate, and novelty) is proposed. The algorithm introduces new metrics and utilizes a genetic algorithm to balance the exploration and exploitation of the population, reducing the likelihood of premature convergence in the particle swarm. Firstly, fitness values, improvement rate, and novelty are used as evaluation metrics for the particles. Each metric is independently employed to select learning examples, which are then stored in separate archives. During each velocity update, particles need to determine the weights of each metric and learn by selecting an example from each archive. Secondly, the algorithm incorporates a genetic algorithm for information exchange among particles. The genetic algorithm introduces cross-swapping and mutation, bringing more randomness to the population and enhancing its global search capability. Finally, eight variants of the PSO algorithm are used as comparative algorithms, and two CEC test suites are employed as benchmark functions in the experiments. The experimental results demonstrate that the FINPSO algorithm outperforms the existing PSO algorithm variants, reaching a state-of-the-art level. Reference | Related Articles | Metrics Select Chaotic Sparrow Search Algorithm with Mixed Multinomial Adaptive Weights DU Yun, ZHOU Zhiqi, JIA Kejin, DING Li, LU Mengyanglin Computer Engineering and Applications    2024, 60 (7): 70-83.   DOI: 10.3778/j.issn.1002-8331.2307-0254 Abstract （74）      PDF（pc） （1361KB）（104）       Knowledge map    Save Sparrow search algorithm has the advantages of simple principle, strong search ability and fast optimization search, but there are shortcomings such as insufficient global search and being easy to fall into local optimization, etc. The paper proposes a chaotic sparrow search algorithm with mixed multinomial adaptive weights to deal with its shortcomings. Improved Circle chaotic mapping is added to improve the diversity of the population; an adaptive weighting strategy is introduced in the explorer to improve the global search ability and expand the search range of the explorer; bubble net predation strategy is introduced in the joiner to improve the whale optimization algorithm, which improves the local search performance of the algorithm and the ability to escape local optima; and combined with the mechanism of the backward learning strategy, the optimal selection of all individuals is made to improve the quality of individuals in each iteration, thereby improving the quality of the algorithm. The quality of individuals is improved to enhance the algorithm’s optimization search efficiency and accuracy. The chaotic sparrow search algorithm with mixed multinomial adaptive weights is compared with four classical baseline algorithms and nine improved sparrow search algorithms on 12 benchmark functions and CEC2022 problems, and the improved algorithms in the paper demonstrate better optimization performance and faster convergence speed. Reference | Related Articles | Metrics Select Nash Equilibrium Optimization Control Method for Formation Problem of Non-Cooperative Swarm Systems CHEN Hongwei, ZHANG Qingjie, GE Yuanzheng, ZHOU Wenhong Computer Engineering and Applications    2024, 60 (7): 84-91.   DOI: 10.3778/j.issn.1002-8331.2309-0140 Abstract （29）      PDF（pc） （689KB）（42）       Knowledge map    Save A distributed non-cooperative optimization control method is proposed for the formation control problem for swarm systems. Firstly, a mathematical model of the formation control for swarm systems and individual cost function are established, and a distributed control protocol framework is designed using consensus strategy. Secondly, the framework for the control protocol with parameters is designed based on Nash equilibrium theory. Then, necessary and sufficient conditions for the time-varying formation of the swarm systems are given, and the stability of the system and parameter selection methods are analyzed using the Lyapunov method. Finally, the effectiveness of the method is verified through the simulation experiment and the improvements in optimization performance and convergence speed. Reference | Related Articles | Metrics Select Dynamic Robust Optimization Algorithm Based on Problem Feature Change Guidance LI Erchao, ZHAO Fengkai Computer Engineering and Applications    2024, 60 (6): 130-146.   DOI: 10.3778/j.issn.1002-8331.2306-0033 Abstract （37）      PDF（pc） （1501KB）（45）       Knowledge map    Save Robust optimization is a new method for solving dynamic optimization problems with the goal of finding solutions that are still acceptable for a long time. Most of the researchers in this field try to find new robust solutions based on their future predicted fitness values. However, the error of predicting the future fitness value is often too large, which makes it difficult to find a better robust solution. To solve this problem, an algorithm framework based on problem feature change guidance (ROOT-PFCG) is proposed for dynamic robust optimization. The change of problem characteristics mainly refers to the objective function value of the solution in the current environment and the floating value of the objective function in the corresponding adjacent environment, and three important indexes are proposed. In the case of prediction and non-prediction, three different fitness decision rules are proposed to select the solution based on the index to ensure that the selected solution is less or not affected by the prediction error, so as to find a better robust solution. On this basis, a new performance evaluation index is proposed. The experimental results on benchmark problems show that the proposed algorithm can better improve the performance of robust solutions, and further analyze the impact of indicators on performance in different situations. On this basis, a better indicator combination method is analyzed. Reference | Related Articles | Metrics Select Adaptive DBSCAN Algorithm Based on Improved Harmony Search MENG Xianghui, WEI Zhaokun, ZHANG Xiaoju, HAN Zhifeng Computer Engineering and Applications    2024, 60 (6): 147-154.   DOI: 10.3778/j.issn.1002-8331.2307-0184 Abstract （44）      PDF（pc） （751KB）（49）       Knowledge map    Save As a classical clustering algorithm, DBSCAN algorithm is widely used in various fields. However, due to the poor adaptability of its parameters, the application effect depends entirely on the setting of parameters. Based on this, an adaptive DBSCAN algorithm based on improved harmony search is proposed to improve the adaptability of DBSCAN algorithm. The algorithm first uses the K-means nearest neighbor algorithm to optimize the initial population, thereby improving the quality of the initial population and providing a high-quality solution for subsequent evolutionary calculations. Secondly, an update operator based on double difference is designed to improve the search ability of the algorithm. Thirdly, two update strategy structures are used to avoid premature convergence of the algorithm, improve the optimization ability of the harmony search algorithm, and comprehensively improve the adaptability of the DBSCAN algorithm. Finally, a variety of datasets are used and comparative experiments are designed to verify the proposed algorithm. Experimental results show that the proposed algorithm has better recognition ability and adaptability. Reference | Related Articles | Metrics Select Expensive High-Dimensional Optimization Algorithm with Three-Stage Adaptive Sampling and Incremental Kriging Assistance GU Qinghua, LIU Sihan, WANG Qian, LUO Jiale, LIU Di Computer Engineering and Applications    2024, 60 (5): 76-87.   DOI: 10.3778/j.issn.1002-8331.2306-0252 Abstract （57）      PDF（pc） （857KB）（81）       Knowledge map    Save Surrogate-assisted evolutionary algorithms have been widely used to solve costly multi-objective optimization problems, but are often restricted to solving problems with low-dimensional decision variables due to the limitations of the surrogate model. In order to solve the expensive high-dimensional multi-objective optimization problem, an improved incremental Kriging-assisted evolutionary algorithm based on the three-stage adaptive sampling strategy is proposed in this paper. The algorithm uses an improved incremental Kriging model to approximate each objective function. This model’s hyperparameters are adaptively updated according to prediction uncertainty, reducing computational complexity while ensuring the model’s accuracy in high dimensions. In addition, the three-stage adaptive sampling strategy is proposed for model management, which divides the sampling process into different optimization stages to tailor the selection of individuals. It is able to guarantee convergence first and improve the convergence speed of the algorithm. To verify the effectiveness of the proposed algorithm, experiments are conducted on two sets of test problems containing various features DTLZ (deb-tiele-laumanns-zitzler), MaF (many-objective function) and path planning real engineering problems compared with state-of-the-art algorithms of the same type. The results show that the algorithm is highly competitive in solving expensive multi-objective optimization problems with high-dimensional decision variables. Reference | Related Articles | Metrics Select Improved Local Mean-Based Pseudo Nearest Neighbor Algorithm LI Yi, ZHANG Desheng, ZHANG Xiao Computer Engineering and Applications    2024, 60 (5): 88-94.   DOI: 10.3778/j.issn.1002-8331.2307-0117 Abstract （75）      PDF（pc） （500KB）（84）       Knowledge map    Save Since the local mean-based pseudo nearest neighbor classification algorithm (LMPNN) is easily influenced by the parameter [k] of neighbors and noise points, an improved local mean-based pseudo nearest neighbor classification algorithm (IPLMPNN) is proposed in this paper. Firstly, the nearest neighbors of the test sample are selected by using the two-layer search rule to improve the quality of selecting neighbors. Secondly, in order to overcome the adverse influence of the subjective weighting method and strengthen the effect of each local mean vector for classification, the attention mechanism is introduced to calculate the distance weighting coefficients. Finally, an improved harmonic mean distance is provided to calculate the weighted multi-harmonic mean distance between a test sample and a local mean vector. Furthermore, the pseudo nearest neighbor points are found by the weighted multi-harmonic mean distances to classify test samples. The proposed algorithm is assessed and compared with relative algorithms for multiple data sets from UCI and KEEL. The experimental results show that IPLMPNN obtains satisfactory classification results. Reference | Related Articles | Metrics Select Feature Selection Algorithm Using Dynamic Relevance Weight XU Huajie, LIU Guanting, ZHANG Pin, QIN Yuanzhuo Computer Engineering and Applications    2024, 60 (4): 89-98.   DOI: 10.3778/j.issn.1002-8331.2306-0230 Abstract （89）      PDF（pc） （2466KB）（90）       Knowledge map    Save When considering the new classification information provided by the candidate features, the features selection algorithm based on mutual information usually ignores that the addition of the candidate features will cause the change of the correlation between the selected features and class labels, which will bring additional new information; in addition, when calculating redundant information, using the form of cumulative summation may lead to the underestimation of the redundancy degree of candidate features. In view of the above problems, the definition of dynamic relevance weight is proposed to more comprehensively consider the new information components brought by candidate features. The improved definition of redundant items is proposed, and the maximum and normalization strategy are adopted to solve the problem of underestimating redundancy. On this basis, the feature selection using dynamic relevance weight (FSDRW) is proposed. Five current mainstream filter-based feature selection algorithms based on mutual information are selected for comparative experiments. Experiments on machine learning test datasets from UCI (University of California Irvine) and ASU (Arizona State University) show that the proposed algorithm works well in classification accuracy and comprehensive performance. Finally, the proposed algorithm is applied to the recognition of microseismic and blasting signals in a reservoir project in Guangxi. The selected features of the algorithm are used for microseismic signal recognition, achieving a classification accuracy of 98.86%, verifying the effectiveness of the algorithm in practical applications. Reference | Related Articles | Metrics Select Hybrid-Strategy Improved Golden Jackal Optimization ZHU Xinglin, WANG Tinghua, LAI Zhiyong Computer Engineering and Applications    2024, 60 (4): 99-112.   DOI: 10.3778/j.issn.1002-8331.2306-0099 Abstract （116）      PDF（pc） （2972KB）（88）       Knowledge map    Save In view of the shortcomings of the golden jackal optimization (GJO) in solving complex optimization problems, such as slow convergence speed and being easy to fall into local optimum, a hybrid-strategy improved golden jackal optimization (IGJO) is proposed. Firstly, when the optimal solution of the algorithm stagnates updating, the Cauchy variation strategy is introduced to enhance the population diversity and improve the global search capability of the algorithm to avoid falling into local optimum. Then, a decision strategy based on weight is proposed to accelerate the convergence of the algorithm by assigning different weights to golden jackal individuals. Experiments with eight benchmark functions and some CEC2017 test functions show that the improved algorithm has better optimization performance and convergence speed. Furthermore, the improved algorithm is applied to optimize the parameters of support vector regression(SVR) model, and its effectiveness is verified by experiments on 5 UCI (University of California, Irvine) datasets. Reference | Related Articles | Metrics Select Bagging Heterogeneous Ensemble Code Smell Detection and Refactoring Priority Division WU Haitao, CAI Yongqi, GAO Jianhua Computer Engineering and Applications    2024, 60 (3): 138-147.   DOI: 10.3778/j.issn.1002-8331.2305-0218 Abstract （38）      PDF（pc） （683KB）（45）       Knowledge map    Save Code smells are symptoms of poor design and implementation choices that may hinder code comprehension and possibly increase change-and fault-proneness. Previous research has focused on the detection of code smells by a single model and failed to provide developers with refactoring recommendations. Aiming at the above problems, a code smell detection and refactoring priority division method based on Bagging heterogeneous ensemble model is proposed. The method exploits the heterogeneity between classifiers to detect three code smells such as Complex Class, Long Method, and Spaghetti Code through the F1 integration strategy, and converts the smell probability by the model into a possibility distribution to provide developers with refactoring recommendations. The experimental results verify and evaluate：(1) the stability and relationship among base classifiers in code smells; (2) the performance of Bagging heterogeneous ensemble model to detect the code smells mentioned above; (3) the validity of the refactoring priority when transforming smell probability into possibility distribution, on 32 versions of 6 open-source systems. The experimental results show that the best base classifier varies with code smell types. At the same time, compared with the base classifiers, Bagging heterogeneous ensemble model has a 0.06~40.51 percentage points increase in F1 and a 0.45~28.37 percentage points increase in AUC. Finally, Cohen’s Kappa test is conducted between refactoring priorities of Bagging heterogeneous ensemble model and six respondents, which are highly consistent. Reference | Related Articles | Metrics Select Research on Granule Vectors Random Forest Classification Algorithm ZHANG Kunbin, CHEN Yumin, WU Keshou, HOU Xianyu Computer Engineering and Applications    2024, 60 (3): 148-156.   DOI: 10.3778/j.issn.1002-8331.2305-0204 Abstract （28）      PDF（pc） （554KB）（37）       Knowledge map    Save Granular computing is a computational paradigm that aligns with human cognitive characteristics, enabling the effective processing of complex data. Random Forest reduces the risk of overfitting associated with individual decision trees by ensembling multiple trees. However, it still faces some overfitting issues. To mitigate overfitting and enhance classification performance, this research introduces the concept of granular vector representation into Random Forest. Granular vectors possess the ability to represent high-dimensional features, thereby capturing more data patterns. The randomness in sample selection aids in controlling overfitting, while using different granular vectors for distinct decision trees enhances model diversity. Experimental results demonstrate that, compared to traditional Random Forest and other enhanced algorithms, the Random Forest algorithm based on granular vector representation exhibits superior generalization capabilities and significantly improves classification accuracy. The correctness and effectiveness of the granular vector-based Random Forest classification algorithm are demonstrated. Reference | Related Articles | Metrics Select Domain Adaptation Method Combined with Discriminant Analysis and Distribution Discrepancy Constraints QIN Jiangwei, TANG Deyu Computer Engineering and Applications    2024, 60 (2): 77-86.   DOI: 10.3778/j.issn.1002-8331.2303-0035 Abstract （47）      PDF（pc） （702KB）（36）       Knowledge map    Save To solve the problem of category structure loss and local feature loss caused by feature transformation based on global distribution adaptation in the process of domain adaptation, a domain adaptation method combined with discriminant analysis and distribution discrepancy constraints is proposed. Firstly, the mean distance metric of the domain data distribution is constructed for domain distribution adaptation; secondly, the class divergence metric is constructed to maintain the class discriminant structure; meanwhile, based on the sample local distribution information, multiple discrepancy weights are designed to constrain the domain distribution discrepancy measure and the class discrimination measure respectively, which realizes the joint optimization of discriminant preservation and locality preservation; finally, based on the feature transformation optimized by the metrics aforementioned, the source domain data and target domain data are projected into the subspace for classification task. During the process of domain adaptation, the proposed method can not only reduce the distribution difference between domains, but also take into account the preservation of category discrimination and local characteristics of data, thus effectively improve the performance of reusage of out-of-domain data. The experimental results on 28 cross-domain classification tasks show that the proposed method significantly outperforms the other related methods in terms of the evaluation metric. Reference | Related Articles | Metrics Select Key Nodes Identification Method Based on Neighborhood K-shell Distribution WU Yali, REN Yuanguang, DONG Ang, ZHOU Aoran, WU Xuejin, ZHENG Shuailong Computer Engineering and Applications    2024, 60 (2): 87-95.   DOI: 10.3778/j.issn.1002-8331.2305-0181 Abstract （58）      PDF（pc） （838KB）（39）       Knowledge map    Save Accurate identification of key nodes in complex networks plays a crucial role in network structure stability and information dissemination. The traditional K-shell algorithm only evaluates the importance of nodes’ location, resulting in low differentiation. Considering the influence of global information and local information of nodes comprehensively, an identification algorithm of key nodes based on the K-shell distribution of neighborhood is proposed. The entropy of the node is defined by the [Ks] value of neighborhood to reflect the K-shell distribution characteristics of the neighbors. The results on 11 network datasets demonstrate the accuracy of proposed method. Reference | Related Articles | Metrics Select Colored Traveling Salesman Problem with Conflict Graph: Model and Algorithms XU Wenqiang, ZHOU Yangming, WANG Zhe Computer Engineering and Applications    2024, 60 (1): 135-144.   DOI: 10.3778/j.issn.1002-8331.2302-0306 Abstract （42）      PDF（pc） （746KB）（37）       Knowledge map    Save The colored traveling salesman problem is an important variant of the multiple traveling salesman problem, which is widely used to model optimization problems in multi-machine engineering systems with overlapping areas. The colored traveling salesman problem is difficult to effectively address in scenarios with conflict, which often arises when two cities cannot be visited by the same salesman. Inspired by existing combinatorial optimization problems with conflict graph, a colored traveling salesman problem with conflict graph is proposed and formally defined in this paper. The colored traveling salesman problem with conflict graph is NP-hard, and the CPLEX exact solver can only optimally solve small-scale problem instances. To solve larger instances, an effective memetic algorithm is proposed in this paper, which integrates an adaptive large neighborhood search (ALNS) to perform local optimization. Compared with the exact algorithms, the proposed memetic algorithm finds better results for 9 out of 20 small-scale instances, and uses less computational time for 18?instances. Compared with heuristic algorithms, the proposed memetic algorithm achieves better results for all 14?medium-scale instances. Finally, an ablation experiment is performed to verify the effectiveness of ALNS used in the proposed memetic algorithm. Reference | Related Articles | Metrics Select Discrete Wild Horse Optimizer for TSP CAI Yanguang, FANG Chuncheng, WU Yanlin, CHEN Huajun Computer Engineering and Applications    2024, 60 (1): 145-153.   DOI: 10.3778/j.issn.1002-8331.2303-0012 Abstract （41）      PDF（pc） （616KB）（55）       Knowledge map    Save A new metaheuristic algorithm called discrete wild horse optimizer (DWHO) is proposed for solving the traveling salesman problem. The minimum position matching value method (MPMV) is applied to discretize and decode the obtained solutions. To improve the algorithm’s search capability, a variable neighborhood search strategy is introduced to enhance its local search capability and accelerate the convergence speed by combining wild horse grazing behavior, mating behavior, exchange and?selection of?leaders. 33 benchmark examples from the standard library TSPLIB are selected for experimentation, and it is compared with two other algorithms, artificial bee colony algorithm with sequence exchange (ABCSS) , discrete spider monkey optimization (DSMO) . The experimental results indicate that the maximum improvement rates of the optimal solutions of DWHO, compared with ABCSS, and DSMO algorithms, are 4.52% and 3.41%, respectively. Moreover, the convergence speed of the discrete horse optimizer in solving TSP is significantly better than the above two algorithms. The results also indicate that the discrete horse optimizer has advantages in terms of solving ability and accuracy. Reference | Related Articles | Metrics Select Joint Triple Embedding for Entity Alignment LI Fengying, LI Jiapeng Computer Engineering and Applications    2023, 59 (24): 70-77.   DOI: 10.3778/j.issn.1002-8331.2302-0087 Abstract （38）      PDF（pc） （757KB）（42）       Knowledge map    Save Entity alignment is the task of identifying entities from different knowledge graphs that point to the same item and is important for KG fusion. Most of the existing approaches are based on graph neural networks that learn the entities embedding by modeling the neighborhood information of the entities. However, graph neural networks-based approaches have difficulty learning triples information in knowledge graphs, the triples information is not sufficiently utilized. In order to solve this problem, an entity alignment model with joint triple embedding is proposed in this paper. The proposed model computes a triple embedding for each entity and then uses this triple embedding for entity alignment. In addition, considering that the relations in the knowledge graphs have different types, in order to exploit these relation types, a relation type-aware calculation method of triple embedding is proposed. Meanwhile, constraints based on relation types are added to this model, to learn the mapping properties of relations. Experiments conducted on three real-world datasets show that this approach outperforms state-of-the-art methods, the effectiveness of the proposed method is verified. Reference | Related Articles | Metrics Select k-NN Density Dominator Component Delegations Based Density Peaks Clustering LYU Hongzhang, YANG Yiyang, YANG Geping, GONG Zhiguo Computer Engineering and Applications    2023, 59 (24): 78-87.   DOI: 10.3778/j.issn.1002-8331.2302-0037 Abstract （65）      PDF（pc） （634KB）（73）       Knowledge map    Save DPC（clustering by fast search and find of density peaks） is inefficient in processing large-scale clustering. [k]（lower case）-NN density dominator component skill can improve such shortcoming. However, representative data points in such skill could have poor ability on representation, which leads to lower clustering quality. The delegation sampling strategy can be used as an improvement on the above issue. The resulting new algorithm not only inherits the efficient characteristics of density dominator component acceleration skill, but also ensures the quality of clustering. This algorithm first constructs [k]-nearest neighbor graph. Then, kernel density is estimated and density dominator component is built. Thirdly, each density dominator component is sampled from its high and low density area, and similarity is computed between each dominator via delegations’ nearest neighbor relationship. Finally, DPC algorithm is conducted with each domain as the data point. The experiments show that the introduction of delegations strategy can improve the performance of the original DPC, and the clustering results are better than some other density clustering algorithms. Reference | Related Articles | Metrics Select Research on Code Detection and DFG Generation in CRCLA Compiler Front-End YANG Chenguang, LI Wei, DU Yiran Computer Engineering and Applications    2023, 59 (23): 63-72.   DOI: 10.3778/j.issn.1002-8331.2301-0102 Abstract （37）      PDF（pc） （725KB）（36）       Knowledge map    Save Aiming at the requirement of automatic mapping of cryptographic algorithms to reconfigurable cryptographic logic array（CRCLA） and providing accurate and concise data flow graph for back-end mapping, a front-end design of data flow graph generation and optimization is proposed. In this design, Flex and Bison are used as the compiler framework. The syntax tree is obtained by lexical and syntactic analysis of the code written by the high-level language C++, and the data flow graph is generated by semantic analysis according to the instruction characteristics of cryptographic algorithm and the hardware structure of CRCLA. There are functions implemented in different ways in the source code, such as S-box replacement and bit replacement, but they can be implemented in CRCLA instead of single operator. In order to solve such problems, this paper designs a graph embedding model based on attention mechanism for detection and recognition, and carries out graph structure replacement. At the same time, function expansion, redundant node elimination and data flow graph layering are used to optimize the data flow graph. The experimental results show that the simplified data flow graph is generated automatically after code identification and optimization, and the performance is improved by about 37% compared with other compiler front-end. Reference | Related Articles | Metrics Select Adaptive Multi-Density Peak Sub-Cluster Fusion Clustering Algorithm CHEN Di, DU Tao, ZHOU Jin, WU Yunzheng, WANG Xingeng Computer Engineering and Applications    2023, 59 (23): 73-85.   DOI: 10.3778/j.issn.1002-8331.2212-0234 Abstract （55）      PDF（pc） （887KB）（74）       Knowledge map    Save The classical density peak clustering algorithm relies too much on the cutoff distance when calculating the local density, it is prone to chain effects when assigning non-central points, and it is difficult to identify the cluster centers with uneven density by manually selecting the cluster center points. To solve the above problems, an adaptive multi-density peak sub-cluster fusion clustering algorithm is proposed. Considering the neighborhood information of the samples, the idea of natural neighbors is introduced into the density peak clustering to realize the self-adaptive calculation of the local density of the sample points. In order to find the centers of the sparse clusters, an automatic cluster center selection strategy is proposed to determine the initial sub-cluster center, and a two-step allocation strategy is used to reduce the probability of the occurrence of chain effects. A K-nearest neighbor similarity measurement criterion is proposed, and the subclusters with high similarity are fused to obtain the final clustering result. Compared with the classical density peak clustering algorithm and its improved algorithms in recent years, the algorithm shows better performance on two-dimensional synthetic datasets and UCI datasets. Reference | Related Articles | Metrics Select Gaussian Process Regression Poisson Multi-Bernoulli Filter with Target Spawning SONG Yingying, SONG Liping Computer Engineering and Applications    2023, 59 (22): 84-91.   DOI: 10.3778/j.issn.1002-8331.2301-0086 Abstract （45）      PDF（pc） （681KB）（54）       Knowledge map    Save In order to solve the problem that the Gamma Gaussian inverse Wishart mixed Poisson multi-Bernoulli （GGIW-PMB） filter cannot estimate the non-elliptic shape target, a method combining the Poisson multi-Bernoulli filter and Gaussian process regression model is proposed to estimate the non-elliptic shape target accurately. In view of the problem that the spawned sub-extended target and its extended shape cannot be extracted effectively, a spawned sub-extended target detection and modeling method is proposed. The spawned sub-extended target is detected by the change of the measurement number, and its state is calculated according to the real scene relationship. Combined with the proposed spawned sub-extended model, Gaussian process regression based Poisson multi-Bernoulli filter with target spawning（GPR-PMBS） is proposed. The simulation results show that compared with GGIW-PMB filter, the proposed algorithm can estimate the non-elliptic shape target more accurately, and can extract the spawned sub-extended target and its shape effectively, which shows good performance in the extended target tracking scenarios with target spawning. Reference | Related Articles | Metrics Select Dung Beetle Optimization Algorithm Guided by Improved Sine Algorithm PAN Jincheng, LI Shaobo, ZHOU Peng, YANG Guilin, LYU Dongchao Computer Engineering and Applications    2023, 59 (22): 92-110.   DOI: 10.3778/j.issn.1002-8331.2305-0021 Abstract （122）      PDF（pc） （922KB）（101）       Knowledge map    Save Dung beetle optimizer（DBO） is an effective meta-heuristic algorithm. Dung beetle optimization algorithm has the characteristics of strong searching ability and fast convergence speed. But at the same time, it also has the disadvantages of unbalanced global exploration and local exploitation ability, easy to fall into local optimization, and weak global search ability. Therefore, an improved DBO algorithm is proposed to solve the global optimization problem, named MSADBO. Inspired by the improved sine algorithm（MSA）, this paper endows dung beetles with global exploration and local development capabilities of MSA to expand their search scope, improve their global search capability, and reduce the possibility of falling into local optimal. Chaotic mapping initialization and mutation operator are added to the perturbation. In order to verify the effectiveness of the proposed MSADBO algorithm, 23 benchmark functions are tested and compared with other well-known meta-heuristic algorithms. The results show that the algorithm has good performance. Finally, in order to further illustrate the practical application potential of MSADBO algorithm, the algorithm is successfully applied to three engineering design problems. Experimental results show that the proposed MSADBO algorithm can deal with practical application problems effectively. Reference | Related Articles | Metrics Select Research on Feature Distribution Distillation Algorithm Under Multiple Tasks GE Haibo, ZHOU Ting, HUANG Chaofeng, LI Qiang Computer Engineering and Applications    2023, 59 (21): 83-90.   DOI: 10.3778/j.issn.1002-8331.2301-0064 Abstract （55）      PDF（pc） （655KB）（60）       Knowledge map    Save The rapid improvement of the performance of convolutional neural networks is at the cost of continuously stacking the number of network layers, as well as multiplying the amount of parameters and storage space. This not only causes problems such as overfitting in the training process of the model, but also is unfavorable for the model to run on resource-constrained embedded devices. Therefore, model compression technology is proposed to solve the above problems. This article mainly studies the feature distillation algorithm in model compression technology. Aiming at the fact that using the feature map of the teacher network to guide the student network in feature distillation can not exercise the feature fitting ability of the student network very well, a distillation algorithm based on feature distribution is proposed. The algorithm uses the concept of conditional mutual information to construct the probability distribution of the feature space of the model, and introduces the maximum mean discrepancy（MMD） to design the MMD loss function to minimize the distance between the feature distribution of the teacher network and the student network. Finally, on the basis of knowledge distillation, the toeplitz matrix is used to share the weight of the student network, which further saves the storage space of the model. In order to verify the feature fitting ability of the student network under the training of the feature distribution distillation algorithm, this paper has carried out experimental verification on three image processing tasks：image classification, target detection and semantic segmentation. The experiment shows that the performance of the proposed algorithm is better than the comparison algorithm in the above three learning tasks and achieves distillation between different network architectures. Reference | Related Articles | Metrics Select Density Peak Clustering Algorithm Optimized by Adaptive Clustering Centers Strategy XU Tongtong, XIE Bin, ZHANG Ximei, ZHANG Chunhao Computer Engineering and Applications    2023, 59 (21): 91-101.   DOI: 10.3778/j.issn.1002-8331.2207-0446 Abstract （61）      PDF（pc） （1889KB）（69）       Knowledge map    Save Density peak clustering（DPC） algorithm is a simple and efficient unsupervised clustering algorithm, which can quickly find the clustering centers to complete clustering. However, the local density is defined by truncation distance without considering the spatial distribution characteristics of sample points. Selecting clustering center points by decision graph has strong artificial subjectivity. When using single allocation strategy, it is easy to cause joint error. Therefore, a density peak clustering algorithm optimized by shared nearest neighbors and adaptive clustering centers strategy（ADPC） is proposed. The shared nearest neighbors are used to define the similarity measure between two points, and the local density is redefined so that it reflects the spatial distribution characteristics of samples. The [γ] value is the product of the sample density[ρ] and relative distance [δ]. The “inflection point” is determined by slope difference between adjacent points. And the [γ] power transformation improves the degree of differentiation between the potential clustering centers and the non-clustering centers. Decision function is used to determine the potential clustering centers. Then, the mean of distance between the potential clustering centers adaptive to determine the real clustering centers. The allocation strategy of non-clustering center points is optimized. Through experiments on UCI and synthetic datasets, the algorithm can select the clustering centers adaptively and improve the clustering performance to some extent. Reference | Related Articles | Metrics Select Random Forest Optimization Algorithm Fusion with Approximate Markov Blanket LUO Jigen, XIONG Lingzhu, DU Jianqiang, NIE Bin, XIONG Wangping, LI Zhiqin Computer Engineering and Applications    2023, 59 (20): 77-84.   DOI: 10.3778/j.issn.1002-8331.2207-0443 Abstract （105）      PDF（pc） （671KB）（103）       Knowledge map    Save The correlation and redundancy of features will directly affect the quality of randomly extracted features of random forests, leading to the weakened convergence of random forests and reducing the accuracy, generalization ability and performance of random forest models. Based on this, this paper proposes a random forest optimization algorithm incorporating approximate Markov blankets, which uses approximate Markov blankets to construct similar feature groups, then draws features from each similar group proportionally to form a feature subset of a single decision tree, and repeats the above process until it reaches the size of the random forest. The algorithm can improve the quality of randomly extracted features by eliminating the correlation and redundancy among features using approximate Markov blankets while ensuring the diversity of random forest features. The experimental comparison on 12 different dimensional UCI datasets shows that the random forest incorporating approximate Markov blanket can eliminate feature correlation and redundancy to a certain extent, improve various evaluation indexes of the model, enhance generalization ability, and be more suitable for high-dimensional data. Reference | Related Articles | Metrics Select Three-Dimensional Composite Chaotic System for Image Encryption YU Wanbo, HUANG Rongrong, WANG Wenjin Computer Engineering and Applications    2023, 59 (20): 85-93.   DOI: 10.3778/j.issn.1002-8331.2301-0104 Abstract （44）      PDF（pc） （2884KB）（58）       Knowledge map    Save Chaos is an important subject of nonlinearity, which has been widely used in information security, physics and economics. It is meaningful to construct new chaotic systems continuously. In view of this situation, a class of three-dimensional trigonometric complex chaotic system containing the number of iteration [t] is constructed. The dynamic behavior of the system is tested and analyzed. By analyzing the bifurcation diagram, Lyapunov, derivative of periodic points and spectral entropy of the system, it is determined that it has strong chaotic characteristics and large chaotic interval. At the same time, it is proved that the chaotic system satisfies the Devaney definition of chaos. Through 0-1 test and NIST test, the chaotic sequence randomly generated by the system is determined to have good quality. Based on this system, an efficient image encryption scheme is constructed, which is simple to implement and fast to encrypt. And further through the simulation test, index coefficients of encryption results are calculated and analyzed, and the results are very close to the theoretical optimal value. Among them, the key space is huge, up to 10252. The results are compared with other methods proposed recently, which can prove that the image encryption algorithm has strong security and reliability. Reference | Related Articles | Metrics Select Important Node Recognition in Hypernetworks Based on Node Propagation Entropy WU Yinghan, TIAN Kuo, LI Mingda, HU Feng Computer Engineering and Applications    2023, 59 (19): 66-74.   DOI: 10.3778/j.issn.1002-8331.2212-0020 Abstract （72）      PDF（pc） （947KB）（48）       Knowledge map    Save It is a basic and challenging task to identify important nodes in hypernetworks, and the related research is of great value for further analysis of network topology and functional characteristics. In order to break through the limitations of the existing important node recognition methods, an important node recognition method based on node propagation entropy is proposed by using hypergraph and information entropy theory. This method takes into account both local and global topology information of nodes, uses the node clustering coefficient and the number of neighbors to represent the local propagation influence of node information, the global influence of the node information is reflected by the shortest path between nodes and K-shell centrality, and fully considers the influence of nodes and their neighbors, the importance of nodes in the network is represented by the size of node propagation entropy finally. Using monotonicity, robustness and SIR propagation model evaluation criteria, compared with other methods on six real networks from different fields, experimental results show that the proposed method can identify the important nodes in the hypernetwork accurately and effectively. Reference | Related Articles | Metrics Select Research on Path Planning Algorithm of Mobile Robot Based on Improved Informed-RRT* JIN Wuxuan, MA Xianghua, ZHAO Jinliang Computer Engineering and Applications    2023, 59 (19): 75-81.   DOI: 10.3778/j.issn.1002-8331.2212-0131 Abstract （119）      PDF（pc） （1075KB）（82）       Knowledge map    Save An improved Informed-RRT* path planning algorithm based on node optimization is proposed to address the characteristics of the current Informed-RRT* algorithm that is slow, poorly purposeful and unsmooth in path planning. The adaptive t-distribution function is introduced to change the distribution probability of random points in different environments to improve the efficiency of the algorithm. Then an elliptic focal bias strategy is used to extend a single bias point to the whole elliptic focal point to make the random tree grow close to the minimum distance between the starting point and the target point, which increases the purposefulness of the algorithm. Finally, the re-election grandparent node strategy is used to reduce redundancy for the whole path, and the symmetric multipole type curve method is used to smooth for the path turns. The improved Informed-RRT* algorithm has higher search efficiency, more purposefulness, and smoother planned paths by comparing the surface with multiple sets of experiments. Reference | Related Articles | Metrics Select Algorithm Compression and Hardware Design Co-Optimization for 3D-CNN QIAN Jiaming, LOU Wenqi, GONG Lei, WANG Chao, ZHOU Xuehai Computer Engineering and Applications    2023, 59 (18): 74-83.   DOI: 10.3778/j.issn.1002-8331.2212-0011 Abstract （68）      PDF（pc） （1216KB）（59）       Knowledge map    Save Recently, 3D convolutional neural networks have attracted significant attention due to their excellent performance in video classification. However, the enormous computing and storage requirements of 3D-CNN inevitably lead to performance and energy efficiency problems during deployment, which severely limits its applicability in scenarios with limited hardware resources. To tackle this challenge, this paper proposes an algorithm-hardware co-design and optimization method called 3D FCirCNN to deploy 3D-CNN efficiently. At the algorithm level, 3D FCirCNN uses block circulant matrix to compress 3D-CNN for the first time and further accelerates it with the fast Fourier transform（FFT）, significantly reducing the computation and storage overhead of the model while maintaining a regular network structure. On this basis, 3D FCirCNN introduces activation, batch normalization, and pooling operations in the frequency domain to eliminate the frequent time domain/frequency domain switching overhead caused by FFT. At the hardware design level, 3D FCirCNN designs a dedicated hardware architecture for the compressed 3D-CNN and makes a series of optimization oriented to hardware resources and memory bandwidth. Experiment on Xilinx ZCU102 FPGA shows that compared with the previous state-of-the-art work, 3D FCirCNN can achieve 16.68 times performance improvement and 16.18 times computational efficiency improvement within an acceptable accuracy loss （<2%）. Reference | Related Articles | Metrics Select Naive Bayes Fusion of Multiple Attribute Weighting and Orthogonal Transformation LIU Haitao, CHEN Chunmei, PANG Zhongxiang, LIANG Zhiqiang, LI Qing Computer Engineering and Applications    2023, 59 (18): 84-97.   DOI: 10.3778/j.issn.1002-8331.2211-0411 Abstract （28）      PDF（pc） （618KB）（30）       Knowledge map    Save Because the Naive Bayes algorithm ignores the correlation of multi-dimensional attributes of data, it leads to great application limitations of classification algorithms. In this paper, an improved Naive Bayes algorithm combining multiple attribute weighting and orthogonal transformation is proposed. Firstly, the contribution degree and related mutual information are used to quantify the correlation between discrete attributes and discrete attribute values to obtain their weights. Then, the orthogonal transformation method is used to eliminate the linear relationship between continuous attributes. Then, the conditional probabilities of the weighted discrete attributes and the continuous attributes after orthogonal transformation are distinguished and calculated to obtain higher classification accuracy and improve the generalization ability of the algorithm. Through the [k]-fold cross-validation on the public data set and the campus card data set, the experimental results show that compared with the latest five improved Naive Bayes algorithms, the accuracy of the proposed algorithm is 7.19~9.94 percentage points higher, and the weighted average F1 value is 6.4~11.64 percentage points higher. Reference | Related Articles | Metrics Select Improved Particle Swarm Optimization Algorithm with Circle Mapping and Sine Cosine Factor XU Fuqiang, ZOU Dexuan, LI Can, LUO Hongyun, ZHANG Meng Computer Engineering and Applications    2023, 59 (17): 80-90.   DOI: 10.3778/j.issn.1002-8331.2211-0290 Abstract （131）      PDF（pc） （5088KB）（103）       Knowledge map    Save Aiming at the problems of slow convergence speed, low accuracy and easy to fall into local optimization in the process of search and optimization of particle swarm algorithms, an improved particle swarm algorithm introducing Circle mapping and sine cosine factor is proposed. The population initialization operation introduced into Circle mapping can obtain a more uniform and diverse initial population, which is conducive to improving the convergence speed and accuracy of the algorithm. The strategy of nonlinearly decreasing inertia weight and introducing the sine cosine factor in the sine cosine algorithm is adopted, which better balances the global exploration ability and local development ability of the algorithm. Inspired by the whale optimization algorithm, a random search strategy with elimination system is proposed and adopted, which enhances the ability of the algorithm to jump out of local optimization and global exploration. The algorithm is simulated on 16 benchmark functions, and compared and analyzed with four particle swarm correlation algorithms and other four swarm intelligent optimization algorithms, which verifies that the proposed improved algorithm has stronger convergence performance and stability. Reference | Related Articles | Metrics Select Fuzzy Clustering Algorithm Combined with Cauchy Distribution and Ant Lion Algorithm WU Chenwen, WANG Shasha, CAO Xuetong Computer Engineering and Applications    2023, 59 (17): 91-98.   DOI: 10.3778/j.issn.1002-8331.2205-0436 Abstract （62）      PDF（pc） （4057KB）（54）       Knowledge map    Save Aiming at the problem that fuzzy clustering depends strongly on the initial clustering centers and easy to fall into local optimal solutions, a fuzzy clustering algorithm combined with Cauchy distribution and ant lion algorithm （CALOFCM） is proposed. The Cauchy distribution function variant ant lion optimization algorithm is introduced, which reduces the binding force of individuals by local extreme points, thus increasing the probability of escaping from the local optimum. The elite ant lions generated by the optimized ant lion algorithm are used as the initial clustering centers of the fuzzy C-means （FCM） algorithm. The comparison experiments of artificial data sets and UCI data sets show that compared with K-means, DBSCAN, FCM, ALOFCM algorithm, the proposed algorithm obtains better clustering effect and has good clustering performance in accuracy, adjusted rand index and normalized mutual information. Reference | Related Articles | Metrics Select Fast Spectral Clustering Based on Anchor Point Extraction with Bisecting k-means LUO Xinglong, HE Xingshi, YANG Xinshe Computer Engineering and Applications    2023, 59 (16): 74-81.   DOI: 10.3778/j.issn.1002-8331.2205-0036 Abstract （72）      PDF（pc） （607KB）（62）       Knowledge map    Save Spectral clustering（SC） has received increasing attention due to its effectiveness in unsupervised learning. However, due to its high computational complexity, it is not suitable for processing large-scale data. In recent years, many anchor points graph-based methods have been proposed to accelerate large-scale spectral clustering. However, the anchor points selected by these methods usually cannot well reflect the information of the original data, which leads to the degradation of clustering performance. To overcome these shortcomings, a fast spectral clustering algorithm based on anchor point extraction with bisecting [k]-means（FCAPBK） is proposed. The method uses bisecting [k]-means to select some representative anchor points from the original data, then constructs a multi-layer kernel-free similarity graph based on anchor points, and constructs a hierarchical bipartite graphs through the similar relationship between the anchor points and the sample. Finally, experiments are carried out on five benchmark datasets, and the results show that the FCAPBK method can obtain good clustering performance in a short time. Reference | Related Articles | Metrics Select Modified Chimp Optimization Algorithm Based on Learning Behavior Strategy JIA Heming, LIN Jiankai, WU Di, LI Shanglong, WEN Changsheng, RAO Honghua Computer Engineering and Applications    2023, 59 (16): 82-92.   DOI: 10.3778/j.issn.1002-8331.2211-0024 Abstract （82）      PDF（pc） （857KB）（80）       Knowledge map    Save Aiming at the problems of slow convergence speed, low optimization accuracy and easy to fall into local optimum of chimp optimization algorithm, a modified chimp optimization algorithm（MChOA） integrating learning behavior strategy is proposed. Firstly, quasi opposition-based learning strategy is used to update the population, increase the diversity and randomness of the population, improve the global search ability of the algorithm, and avoid the algorithm from falling into local optimum. Then, based on the learning behavior strategy of chimps, the individual position of chimps is updated by randomly selecting the “imitation learning” operator or the “emotional induction” operator to enhance the local development ability of the algorithm and speed up the convergence of the algorithm. 16 benchmark functions and 12 CEC2014 are selected for simulation experiment test, the results show that MChOA has higher precision and better optimization performance than traditional ChOA. Finally, the solution of two engineering design problems proves that MChOA has high practical application value in practical engineering problems. Reference | Related Articles | Metrics Select Adaptive Neighbor Local Ratio Sum Linear Discriminant Analysis ZHANG Jiale, LIN Haoshen, ZHOU Keyi, SUN Bo, YANG Xiaojun Computer Engineering and Applications    2023, 59 (15): 115-122.   DOI: 10.3778/j.issn.1002-8331.2204-0348 Abstract （56）      PDF（pc） （594KB）（39）       Knowledge map    Save In machine learning and pattern recognition, dimension reduction can significantly improve the discriminative performance and efficiency of the classifier. Ratio sum（RS）is a completely new variant of linear discriminant analysis（LDA） that tries to optimize the projection matrix in every dimension. However,  the RS does not take into account the local geometry of the data,  which may lead to the failure to obtain the optimal solution. To overcome this disadvantage of RS,  an adaptive neighbor local ratio sum linear discriminant analysis algorithm is proposed. The algorithm uses the adaptive neighbor composition method to construct the neighbor matrix,  retains the local geometric structure of the data to complete the construction between and within the data matrix,  so as to better find the optimal representation of the data. And the method adopts the effective kernel-free parameter neighborhood allocation strategy to construct the neighbor matrix,  to avoid the need to adjust the heat core parameters. Finally, comparative experiments on the UCI dataset verifies the effectiveness of the algorithm. Reference | Related Articles | Metrics Select Density Peak Clustering Algorithm Based on Space Vector Search MA Zhenming, AN Junxiu Computer Engineering and Applications    2023, 59 (15): 123-131.   DOI: 10.3778/j.issn.1002-8331.2204-0179 Abstract （53）      PDF（pc） （874KB）（40）       Knowledge map    Save A density peak clustering（VS-DPC） algorithm based on spatial vector search is proposed to address the problem of excessive time overhead due to the construction of the similarity matrix between global sample points in the density peak clustering（DPC） algorithm. Firstly, the data points are mapped into a spatial vector starting from the origin in an n-dimensional orthogonal coordinate system, and the modulus of the vector and the angle between it and the positive direction of the unified coordinate axis are calculated. Secondly, the similarity vector is searched using the truncation distance and truncation mapping angle to determine the similarity range. Finally, the similarity vector is used to determine the effective density points thus constructing a sparse similarity matrix to reduce the time complexity. The experimental results on seven real datasets and seven artificial datasets with complex shapes in the UCI database show that the proposed VS-DPC algorithm maintains the clustering accuracy of DPC and reduces the time overhead by about 60% compared to the DPC algorithm. And compared with CDPC and GDPC, two algorithms to improve the efficiency of DPC clustering, the algorithm has fewer parameters and improves the clustering accuracy and time by 6 and 18 percentage points on average, respectively. Reference | Related Articles | Metrics Select Hybrid Algorithm of Bird Swarm Algorithm and Arithmetic Optimization Algorithm Based on Cauchy Mutation LU Mengdie, LU Haiyan, HOU Xinyu, ZHAO Jinjin, XU Jie Computer Engineering and Applications    2023, 59 (14): 62-75.   DOI: 10.3778/j.issn.1002-8331.2208-0463 Abstract （105）      PDF（pc） （767KB）（96）       Knowledge map    Save Aiming at the problems that the bird swarm algorithm has insufficient population diversity in the initial iterations, slow convergence speed in the later iterations and the tendency to fall into local optimum solution, a hybrid algorithm of bird swarm algorithm and arithmetic optimization algorithm based on Cauchy mutation（HBSAAOA） is proposed. Firstly, the location of producers in BSA is updated by using the high distribution of multiplication and division operators in the arithmetic optimization algorithm to improve the diversity of population and hence to enhance the global search ability. Then, a random search strategy and a Cauchy mutation strategy are introduced to generate candidate solutions, which will disturb the local exploitation in the later stage to enhance the ability of the algorithm jumping out of local optimum solution and to improve the convergence speed. Finally, the greedy strategy is used to select the best individual to replace the poor and thereby to improve the quality of the solution. Through the simulation experiments of 23 classical test functions and some CEC2014 benchmark functions, and applying HBSAAOA to two engineering application problems, the results show that the improved strategies are effective, and the improved algorithm has faster convergence speed, higher optimization accuracy and better robustness. Reference | Related Articles | Metrics Select Flower Pollination Algorithm Combining Lens Imaging and Traction Mutation and Its Application LI Dahai, WU Zhaoqian, WANG Zhendong Computer Engineering and Applications    2023, 59 (14): 76-85.   DOI: 10.3778/j.issn.1002-8331.2209-0332 Abstract （49）      PDF（pc） （862KB）（63）       Knowledge map    Save Due to defects of relatively slow convergence and being easily trapped in local optimal of flower pollination algorithm（FPA）, an enhanced lens learning and traction mutation based flower pollination algorithm（LMFPA） is proposed in this paper. First, LMFPA applies a modified lens learning mechanism to improve the distribution of population. Second, LMFPA uses the observation factor based traction mutation strategy to further accelerate the convergence and increase the probability to jumping out of local optimal. 12 benchmark functions from CEC2013 are selected as testbed to evaluate performance of LMFPA with original FPA algorithm and another 3 improved FPA algorithms, including TMFPA （T-distribution mutation-based flower pollination algorithm）, t-GSSA（improved sparrow search algorithm based on adaptive t-distribution and golden sine and its application） and PCSPSO（particle compaction and scheduling based particle swarm optimization）. Experimental result shows that LMFPA can achieve superior performance both in convergence speed and accuracy. At last, LMFPA is also used to solve 3D path planning for UAVs. Experimental result illustrates that LMFPA can also find better 3D paths for UAVs. Reference | Related Articles | Metrics Select Formation Control of Non-Holonomic Wheeled Mobile Robot in Finite Time QIAO Lei, LI Zonggang, DU Yajiang Computer Engineering and Applications    2023, 59 (13): 74-81.   DOI: 10.3778/j.issn.1002-8331.2207-0396 Abstract （58）      PDF（pc） （746KB）（45）       Knowledge map    Save Keeping the formation shape intact is an important prerequisite for multi-robot formation to work normally in various environments. To solve the problem, this paper proposes a distance-based formation control strategy for multiple nonholonomic mobile robots in leader-follow framework in finite time, in which the target formation is assumed to be minimal and infinitesimally rigid. Since only part of the followers can directly obtain the leader's information, a distributed velocity estimator is designed for the followers to estimate the leader’s velocity. Meanwhile, the leader and the follower’s velocity can reach the consensus within a finite time. Besides of this, a formation control law is proposed to achieve the desired formation shape. In addition, it is proved that both distributed velocity estimator and formation controller can play a role in the local coordinate. Finally, the feasibility of the proposed algorithm is verified by simulation analysis. Reference | Related Articles | Metrics Select Hybrid Variable Neighborhood Tabu Search Algorithm for Vehicle Routing Problem with Hard Time Window HE Qi, GUAN Lihe, CUI Huanhuan Computer Engineering and Applications    2023, 59 (13): 82-91.   DOI: 10.3778/j.issn.1002-8331.2208-0431 Abstract （84）      PDF（pc） （751KB）（85）       Knowledge map    Save To find a high-quality approximate solution to the vehicle routing optimization problem with hard time window, a dual objective nonlinear optimization model is established to minimize the number of vehicles and the total driving distance, because the insufficient consideration of time window constraints in the existing mathematical models, and a hybrid variable neighborhood tabu search algorithm is proposed to solve it. On the one hand, an improved saving algorithm is used to generate the initial solution. And three deletion operations and one insertion operation are designed to optimize the number of vehicles in the initial solution, so as to provide an excellent initial solution for the subsequent tabu search. On the other hand, four neighborhood construction operators are designed in the tabu search process. The flexible storage structure, tabu criterion for avoiding repeated searches and amnesty criterion for enhancing diversity search can effectively get rid of local optimal solution, and finally achieve global optimization. Finally, the experimental analysis is carried out on 56 benchmark examples of Solomon and 18 benchmark examples of Homberger. This algorithm has good convergence and stability, and its solutions are better than the two similar search algorithms in the literature. Moreover, the total driving distance is lower than the currently known best solution on 42 examples. Reference | Related Articles | Metrics page Page 1 of 3 Total 98 records First page Prev page Next page Last page