胡杰, 徐泽俊, 岳肖, 李昊, 袁春, 宾洋. 质子交换膜燃料电池性能优化控制[J]. 中国电机工程学报, 2024, 44(1): 395-406. DOI: 10.13334/j.0258-8013.pcsee.221889
引用本文: 胡杰, 徐泽俊, 岳肖, 李昊, 袁春, 宾洋. 质子交换膜燃料电池性能优化控制[J]. 中国电机工程学报, 2024, 44(1): 395-406. DOI: 10.13334/j.0258-8013.pcsee.221889
HU Jie, XU Zejun, YUE Xiao, LI Hao, YUAN Chun, BIN Yang. Performance Optimization Control of Proton Exchange Membrane Fuel Cells[J]. Proceedings of the CSEE, 2024, 44(1): 395-406. DOI: 10.13334/j.0258-8013.pcsee.221889
Citation: HU Jie, XU Zejun, YUE Xiao, LI Hao, YUAN Chun, BIN Yang. Performance Optimization Control of Proton Exchange Membrane Fuel Cells[J]. Proceedings of the CSEE, 2024, 44(1): 395-406. DOI: 10.13334/j.0258-8013.pcsee.221889

质子交换膜燃料电池性能优化控制

Performance Optimization Control of Proton Exchange Membrane Fuel Cells

  • 摘要: 质子交换膜燃料电池(proton exchange membrane fuel cell,PEMFC)的热管理,是影响PEMFC安全、稳定、高效运行的核心关键因素。该文在分析PEMFC热管理子系统的结构,以及关键组成部件(包括空气压缩机、冷却液循环泵和冷却风扇)的工作原理基础上,首先搭建耦合热管理子系统的PEMFC非线性仿真模型,进一步将其简化处理得到用于控制的线性模型;在此基础上,设计基于卡尔曼滤波的状态观测器,以实现对状态和不可测量扰动的估计,从而消除模型稳态误差;随后设计目标计算器,以净输出功率跟踪误差最小化和系统效率最大化为目标,对PEMFC系统工作状态进行优化;进而基于线性控制模型,设计无偏显式模型预测控制(offset-free explicit model predictive control,offset-free EMPC),对优化工作状态进行跟踪控制;最后,搭建一个软件在环(software in the loop,SiL)测试平台,通过与传统MPC控制策略的仿真结果对比可得,无偏EMPC控制策略可以显著将净输出功率的平均跟踪误差绝对值从344.29W减小到82.19W,系统效率从40.32%提升到46.56%,算法平均单步计算耗时从26ms降低到11.8ms,并且可以约束PEMFC系统运行在稳定和安全的工况。

     

    Abstract: The thermal management of proton exchange membrane fuel cell (PEMFC) is the key factor affecting the safe, stable and high efficient of PEMFC. Based on the structure analysis of the PEMFC thermal management subsystem and its operating principle of key components (including air compressor, coolant circulation pump and cooling fan), a PEMFC simulation model coupling with the thermal management subsystem is created firstly, and it is linearized further to obtain a simplified control orientated linear model. Then, a state observer based on Kalman filtering is designed to estimate the state and non-measurable perturbations, for the purpose of eliminating static errors. To minimize the net output power tracking error and maximize the system efficiency, a target calculator is designed to optimize the state variables of the PEMFC system. Consequently, an offset-free explicit model predictive control (offset-free EMPC) is proposed based on the control orientated linear model to track the optimal state variables. Finally, a kind of software in the loop (SiL) test platform is created. By comparing with the traditional MPC strategy, the SiL simulation results demonstrate that the offset-free EMPC is able to reduce average tracking error of the net output power from 344.29W to 82.19W, the system efficiency is increased from 40.32% to 46.56%, and the strategy single-step average time consumption is reduced from 26ms to 11.8ms. Furthermore, the PEMFC system can be constrained over a stable and safe operating conditions.

     

/

返回文章
返回