Computer Engineering and Applications ›› 2007, Vol. 43 ›› Issue (19): 240-244.

• 工程与应用 • Previous Articles     Next Articles

Application of cellular ant algorithm optimization controlling of fuel cell(PEMFC) engine

WANG Zhou-mian1,2,XU Xiao-ming3,MA Liang1   

  1. 1.Managememt College,University of Shanghai for Science and Technology,Shanghai 200093,China
    2.Armed Police Force Political Academy of Shanghai,Shanghai 200435,China
    3.Automobile College,Tongji University,Shanghai 200092,China
  • Received:1900-01-01 Revised:1900-01-01 Online:2007-07-01 Published:2007-07-01
  • Contact: WANG Zhou-mian

燃料电池发动机优化控制建模与元胞蚂蚁算法

王周缅1,2,徐晓明3,马 良1   

  1. 1.上海理工大学 管理学院,上海 200093
    2.武警上海政治学院,上海 200435
    3.同济大学 汽车学院,上海 200092
  • 通讯作者: 王周缅

Abstract: In this paper,NN is used to model the controlling system of fuel cell engine.And to shorten the temperature rising time and advance the output power of Fuel Cell Engine,a new evolutional algorithm is applied to the optimization controlling of Fuel Cell Engine.Ant Colony Optimization(ACO) is a new optimization algorithm,and it has many good features.So it is widely applied to complication combinatorial optimization problems.But there is much deficiency.Specially,its searching speed is slow.And it will easily get into local optimization.This paper gives a new optimization algorithm which combines the principle of cellular automata with ant algorithm.The algorithm raises its searching speed with evolving the living state of every node.Simulation in Fuel Cell(PEMFC) Engine cold start controlling shows that the result is better than ant algorithm.

摘要: 以某燃料电池发动机为原型,将神经网络辨识方法应用到其非线性系统的建模。在此基础上,将元胞蚂蚁算法应用于燃料电池发动机优化控制问题,以低温启动时暖机时间(升温至正常工作温度范围)最短和提高燃料电池发动机的输出功率为优化控制目标。蚂蚁算法是一种新型的进化算法,它具有许多优良的性质,因此被广泛用于求解各类组合优化问题。但基本的蚂蚁算法有许多不足,比如它搜索速度慢。将元胞自动机与蚂蚁算法结合,通过对结点的生死状态演变,加快了算法的寻优速度,并最大可能地避免陷入局部最优。最后通过对模型的测试,采用元胞蚂蚁算法优化后的控制方法基本达到了上述目标,证明了建模的合理性的同时,显示出了该文优化控制方法的优越性。

关键词: 元胞蚂蚁算法, 燃料电池发动机, 神经网络建模, 优化控制