摘要
针对传统模拟退火算法在电动汽车制动控制中还存在容易陷入局部最优等问题,本文提出一种基于电动汽车制动控制的遗传算子优化模糊退火算法,首先对遗传算法的遗传算子进行优化,对交叉算子进行算术交叉,自适应原算法的变异算子,提高了原算法的全局寻优能力,然后采用蚁群算法对传统算法进行种群的关联优化、信息素更新优化和种群多样性优化,接着将改进的遗传算法和模拟退火算法进行融合,用复制操作、交叉退火操作和变异退火操作提高模拟退火算法的全局寻优能力。通过仿真实验表明,本文提出的基于电动汽车制动控制的蚁群遗传算法优化模糊退火算法收敛性更好,并且能耗更低。
According to the traditional simulated annealing algorithm is easily to fall into local optimum in the electric vehicle brake control, we propose a fuzzy annealing algorithm optimized by genetic operators based on electric vehicle brake control. First, optimize the genetic operators of genetic algorithm, do arithmetic crossover for crossover operator, and adapt to mutation operator of the original algorithm, improve the global optimization of the original algorithm. Then ant colony algorithm is adopted to optimize the associated populations, pheromone updating and population diversity of the traditional algorithms, and then the improved genetic algorithm and simulated annealing algorithm are mixed, improve the global search capability of the simulated annealing algorithm with the copy operation, cross annealing operation and mutation operation. Simulation results show that the proposed fuzzy annealing algorithm optimized by genetic operators based on electric vehicle brake control has better convergence, and lower power consumption.
出处
《中国农机化学报》
2016年第2期205-209,共5页
Journal of Chinese Agricultural Mechanization
基金
江苏省自然科学基金面上项目(BK20131221)
关键词
电动汽车
制动控制
全局寻优
遗传退火算法
蚁群算法
electric vehicles
brake control
global optimization
genetic annealing algorithm
ant colony algorithm
