Modeling And Predictive Control Of Proton Exchange Membrane Fuel Cells

Proton Exchange Membrane Fuel Cell(PEMFC)is a clean and efficient power generation device that can convert chemical energy into electrical energy.Nowadays,hydrogen energy develops rapidly,and PEMFC,as a kind of hydrogen energy carrier,also gets more and more attention.Compared with other traditional energy sources,PEMFC has many advantages,such as non-pollution,high efficiency,low operating temperature,fast starting speed,and so on.It has huge potential and broad application prospects.Nowadays,research on PEMFC has involved numerous directions,among which the emphasis on its dynamic output characteristics is becoming increasingly significant.PEMFC is a complex system with multiple inputs,multiple outputs,strong coupling,and nonlinearity.In order to ensure that the fuel cell system can respond quickly under load changes,and solve the problem of difficult to control the dynamic characteristics of the system output,so that the system can operate continuously and stably under various working conditions,research on modeling and control algorithms is necessary.There have been many studies on mechanism models and identification models,but most of them have not been organically combined for corresponding research.This article will establish a system level mechanism model based on the electrochemical and other internal mechanisms of PEMFC.In order to simplify the difficulty of controller development,subspace identification is carried out based on the input and output data of the mechanism model,and then relevant control strategy research is carried out based on this identification model.The main work done includes:(1)Using the internal mechanism equations such as electrochemistry and thermodynamics of PEMFC,a 40 k W level Simulink simulation model including fuel cell stacks and peripheral auxiliary components is established,and the operating mechanism of the fuel cell power generation process is studied through the model,providing data support for subsequent identification modeling.(2)Aiming at the complex internal working conditions of the PEMFC system,the numerous mechanism modeling parameters that are difficult to obtain,and this kind of model is not suitable for control oriented research,this paper proposes to use subspace identification algorithms as a data driven modeling method.Select appropriate modeling variables,combine the collected input and output data,and identify the required state space model,laying a foundation for subsequent control strategy research.(3)In order to enable PEMFC to output safely and efficiently according to requirements,based on the identified state space model,this paper designs an MPC control strategy for net output power,and then compares the MPC control strategy with the effectiveness of traditional PID control.Experiments show that MPC control can play a better role in the output control of the system than traditional PID control,and the dynamic response ability of the system is significantly improved,The output power is more stable,and the service life of the stack will be correspondingly extended.