混合动力汽车能量管理全局优化算法仿真

Simulation on Global Optimization Algorithm for Energy Management of Hybrid Electric Vehicle

串联混合动力汽车(SHEV)的能量管理是协调分配发动机和电池两个动力源的功率输出,以达到降低油耗的目的。猴群算法作为一种群智能全局优化算法可以实现汽车油耗的全局最优,但要达到指定精度需要的计算时间过长。针对这一问题,提出了基于改进猴群算法的SHEV能量管理策略,优化发动机功率的可行域,降低算法寻优的错误率,采用Logistic混沌函数对猴群进行初始化,使初始猴群在可行域内均匀分布,在爬过程采用先大步长后小步长爬山的方式,加快在爬山前期的优化速度,并在望过程后加入模式搜索算法对位置最佳猴子的周围进行精细搜索,提高算法的局部搜索能力。采用matlab编程,对10次仿真结果求平均值,结果表明改进算法在得到油耗全局最优解的同时拥有更快的收敛速度。

Series hybrid electric vehicle （SHEV） energy management is coordinated distribution of the power of two energy sources, engine and battery, to decrease fuel consumption. As a swarm intelligence global optimization algorithm, monkey algorithm can achieve global optimum of fuel consumption, but it spends too much time achieving the specified precision. Directing against the problem, an improved monkey algorithm was proposed to manage energy of SHEV. The feasible region of the power of engine was optimized to reduce the error rate of algorithm. The chaos function logistic was used to initialize monkeys which makes initial monkeys well- distributed, utilized climbing process with long step first and short step later to accelerate the optimization speed of climbing, and added pattern search algorithm after looking process to finely search around the monkey with the best position. Though Matlab programming, ten simulation results were averaged. The results show that the improved algorithm has faster convergence speed when getting the global optimization of fuel consumption.