关键词: 算术优化算法, 非线性控制参数, sigmoid函数, 小孔成像学习, CEC2014

Abstract: In order to improve the global exploration and local development performance of the arithmetic optimization algorithm, an arithmetic optimization algorithm（IX-AOA） based on the non-linear aperture imaging mechanism and the optimal position-guided search strategy with fine-tuning function is proposed. The adaptive adjustment of the arithmetic acceleration optimizer factor is a vertical update strategy based on the exploration and development purpose of the balanced algorithm, adjusts the algorithm exploration and development time allocation through the hyperbolic nature, and then realizes the algorithm’s self-update mechanism to achieve the effect of balancing algorithm exploration and development. It uses sigmoid activation in the optimal individual, the weight coefficient of the function adjusts the reasonable control mechanism of the search and development of the offspring position to achieve the goal of enriching high-quality candidate solutions. It introduces the principle of nonlinear aperture imaging to improve the search mechanism to optimize the stagnation of optimization caused by the phenomenon of coverage between candidate solutions. Finally, 8 benchmark functions and some CEC2014 and Wilcoxon rank sum tests are used to verify the superiority of the algorithm function optimization.

Key words: arithmetic optimization algorithm, non linear control parameters, sigmoid function, small hole imaging learning, CEC2014