基于HHO-FA的PEMFC电堆辨识建模

IDENTIFICATION MODELING OF PEMFC STACK BASED ON HHO-FA

为解决质子交换膜燃料电池(PEMFC)模型参数难以确定的问题,该文提出一种基于哈里斯鹰算法(HHO)和萤火虫算法(FA)联合的优化算法,即HHO-FA算法,用于PEMFC模型的参数辨识。为提高PEMFC建模精确度,HHO-FA保留HHO中搜索效率和精度较高的全局搜索过程,局部寻优过程结合具有群体寻优特征的FA算法,同时优化负责全局搜索和局部搜索切换的转换因子,加入惯性权重因子,优化算法结构。该文使用燃料电池的商业仿真工具箱Thermolib获取算例数据,并通过与粒子群算法(PSO)、HHO算法、蚁群算法(ACO)和FA算法对比分析,对HHO-FA的PEMFC参数辨识性能进行研究。仿真结果表明,相较于PSO、HHO、ACO和FA,HHO-FA的辨识精确度和收敛效率均最高,证实所提出HHO-FA算法在PEMFC模型参数辨识方面的突出性能。

To solve the problem that the parameters of the model of proton exchange membrane fuel cell(PEMFC)are difficult to determine,a novel joint optimization algorithm is proposed in this paper,integrating the Harris hawks optimizer(HHO)and firefly algorithm(FA).The HHO-FA algorithm is employed to tackle the parameter identification problem of PEMFC.To enhance the modeling accuracy of PEMFC,HHO-FA retains global exploitation with high search efficiency and accuracy in HHO.The local exploitation is combined with the FA with the characteristics of group optimization.At the same time,the conversion factor responsible for switching between global exploration and local exploitation is optimized,and the inertia weight factor is added to optimize the algorithm structure.Sample data is obtained using Thermolib,a commercial simulation toolbox based on fuel cells.The performance of HHO-FA for PEMFC parameter identification is evaluated against particle swarm optimization(PSO),HHO algorithm,ant colony optimization(ACO)and FA algorithm.The simulation results show that,compared with PSO,HHO,ACO and FA,HHO-FA has the highest identification accuracy and convergence efficiency,which confirms the outstanding performance of the proposed HHO-FA algorithm in PEMFC parameter identification.