基于AEO-MRFO的固体氧化物燃料电池参数辨识

Parameter Identification of Solid Oxide Fuel Cell Based on AEO-MRFO

提出一种基于人工生态系统优化(artificial ecosystem-based optimization,AEO)算法与蝠鲼觅食优化(manta ray foraging optimization,MRFO)算法的协调优化算法,即AEO-MRFO协调优化器(AEO-MRFO coordinating optimizer,EMCO),用于各种固体氧化物燃料电池(solid oxide fuel cell,SOFC)的参数辨识。为提高SOFC参数辨识的精确度与稳定性,EMCO舍弃MRFO气旋觅食算子中随机性过强的搜索操作,并随迭代过程动态协调AEO分解算子和经过改进的MRFO翻滚觅食算子,合理平衡局部探索(local exploitation)和全局搜索(global exploration)。通过4个算例对EMCO的SOFC参数辨识性能进行研究,即荷兰能源研究中心和波兰CEREL公司各自生产的2种SOFC单体电池测试数据和蒙大拿州立大学的5kW SOFC电池堆栈在2个不同运行条件下的实验数据。此外,还研究了SOFC堆栈温度及压强变化对参数辨识精度的影响。仿真结果表明,与蚁群优化(ant colony optimization,ALO)算法、均衡优化器(equilibrium optimizer,EO)、灰狼优化(grey wolf optimization,GWO)算法、堆栅优化器(heap-based optimizer,HBO)、粒子群优化(particle swarm optimization,PSO)算法、AEO算法和MRFO算法相比,EMCO均能快速、精确、稳定地实现SOFC参数辨识。

A coordinating optimization method based on artificial ecosystem-based optimization(AEO) algorithm and manta ray foraging optimization(MRFO) algorithm is proposed in this paper,i.e.AEO-MRFO coordinating optimizer(EMCO),which is utilized for parameter identification of solid oxide fuel cell(SOFC).To enhance the precision and the stability of parameter identification of SOFC,EMCO discards strong random search operations of cyclone foraging operator of MRFO,and dynamically coordinates decomposition operator of AEO with revised somersault foraging operator of MRFO,which can properly balance the local exploitation and the global exploration in the optimization process.Four case studies are carried out to investigate the performance of EMCO for parameter identification of SOFC is investigated through the studies in four cases,including two test data sets of single SOFCs developed by Energy Research Centre of Netherlands and CEREL of Poland respectively,as well as two experimental data sets of a 5 kW SOFC stack at Montana State University under two different operating conditions.Besides,the impacts for the precision of the parameter identification have also been studied under variable temperatures and pressures of SOFC stack.The simulation results indicate that,compared with ant colony optimization(ACO) algorithm,equilibrium optimizer(EO),grey wolf optimization(GWO)algorithm,heap-based optimizer(HBO),particle swarm optimization(PSO) algorithm,AEO algorithm and MRFO algorithm,EMCO has the superiority of rapid,accuracy and stability for SOFC parameter identification.