摘要
针对获得准确蓄电池模型的问题,提出一种基于改进蜻蜓算法的铅酸蓄电池参数辨识方法。原始蜻蜓算法存在收敛效率低和易陷入局部最优的缺陷。利用Tent混沌映射具有均匀性和有序性的特点,提高初始种群的多样性;利用精英反向群策略能充分利用优秀个体的特点,扩大种群的搜索空间;引入S型函数改进蜻蜓算法的步长更新权重,S型函数具有前期变化速度较快、后期趋于平稳的特点,可提高算法由全局搜索向局部开发过渡的平衡性,且在各阶段均保持较高的寻优效率。标准测试函数验证了改进算法的效果。利用改进蜻蜓算法对铅酸蓄电池的三阶Thevenin电路模型进行参数辨识,并与蜻蜓算法和带有繁殖机制的粒子群算法进行比较。辨识结果表明,改进蜻蜓算法在蓄电池参数辨识方面具有更高的精度和稳定性。
To obtain an accurate lead-acid battery model,a parameter identification method for battery based on improved dragonfly algorithm is proposed.The classical dragonfly algorithm has the shortcomings,such as low convergence efficiency and being easy to trap into local optimum.Tent map was used to improve the diversity of the initial population based on the uniform and ordered characteristics of tent chaotic map.Elite reverse group strategy was applied to expand the search space of the population for the characteristics making full use of excellent individuals.S-type function was introduced to improve the step update weight of dragonfly algorithm.S-type function has the characteristics of fast change in the early stage and tends to be stable in the later stage,which can improve the balance of the transition from global search to local development,and maintain high optimization efficiency in each stage.The standard test functions verify the effect of the improved algorithm.Finally,the third-order Thevenin circuit model was identified by the improved dragonfly algorithm,and compared with the dragonfly algorithm and the particle swarm algorithm with reproductive mechanism.The identification results show that the improved dragonfly algorithm has higher precision and stability in nonlinear system identification.
作者
吴忠强
赵德隆
王云青
刘重阳
WU Zhong-qiang;ZHAO De-long;WANG Yun-qing;LIU Chong-yang(Key Lab of Industrial Computer Control Engineering of Hebei Province,Yanshan University,Qinhuangdao 066004,China)
出处
《电机与控制学报》
EI
CSCD
北大核心
2020年第12期152-160,共9页
Electric Machines and Control
基金
国家自然科学基金重点资助项目(U1260203)
河北省自然科学基金资助项目(F2020203014)。
关键词
电池
算法
优化
平衡
辨识
非线性系统
batteries
algorithms
optimization
balancing
identification
nonlinear systems
