PEMFC Identification Based on a Fractional-Order Hammerstein State-Space Model with ADE-BH Optimization  

作　　者：Qin Hao Qi Zhidong Ye Weiqin Sun Chengshuo 

机构地区：[1]Nanjing University of Science and Technology,Nanjing 210094,China

出　　处：《China Petroleum Processing & Petrochemical Technology》2023年第2期155-164,共10页中国炼油与石油化工（英文版）

基　　金：This project is supported by the Postgraduate Research&Practice Innovation Program of Jiangsu Province(SJCX22_0124);the National Natural Science Foundation of China(NO.61374153).

摘　　要：Considering the fractional-order and nonlinear characteristics of proton exchange membrane fuel cells(PEMFC),a fractional-order subspace identification method based on the ADE-BH optimization algorithm is proposed to establish a fractional-order Hammerstein state-space model of PEMFCs.Herein,a Hammerstein model is constructed by connecting a linear module and a nonlinear module in series to precisely depict the nonlinear property of the PEMFC.During the modeling process,fractional-order theory is combined with subspace identification,and a Poisson filter is adopted to enable multi-order derivability of the data.A variable memory method is introduced to reduce computation time without losing precision.Additionally,to improve the optimization accuracy and avoid obtaining locally optimum solutions,a novel ADEBH algorithm is employed to optimize the unknown parameters in the identification method.In this algorithm,the Euclidean distance serves as the theoretical basis for updating the target vector in the absorption-generation operation of the black hole(BH)algorithm.Finally,simulations demonstrate that the proposed model has small output error and high accuracy,indicating that the model can accurately describe the electrical characteristics of the PEMFC process.

关 键 词：PEMFC Hammerstein model Fractional subspace identification ADE-BH optimization 

分 类 号：TM911.4[电气工程—电力电子与电力传动]