Modeling And Control Strategy Of Fuel Cell System Under Multiple Operating Conditions

The popularity of vehicles and other means of transportation has accelerated the formation of the greenhouse effect.The development of new energy vehicles such as hydrogen vehicles is an important approach to carbon neutrality.Proton exchange membrane fuel cell(PEMFC)has the advantages of high energy conversion efficiency and pollution-free products,it has extensive application prospects in the field of hydrogen vehicles.Countries in the world are actively involved in promoting the industrialization of fuel cell systems however,the lifetime of fuel cell systems currently restricts its largescale promotion.The control strategies of the fuel cell system under multiple operating conditions are the key factors that influence its lifetime.To prolong the lifetime of the fuel cell system,this dissertation will analyze the key parameters that affect the system lifetime through the dynamic experiments of multiple operating conditions including startup/shutdown and working,and develops dynamic models for two operating conditions respectively.The coupling mechanism between the capacitance effect and carbon corrosion in the stack is studied,and fast and stable control strategies for temperature and oxygen excess ratio are proposed.The main contents and contributions are as follows:(1)A fuel cell system experimental platform is built,the key parameters that affect the system lifetime are studied and the optimization methodology of each parameter is analyzed.The platform provides the experimental fundament for the development and verification of the system model.The four key factors that influence system lifetime are the fuel cell degradation rate during the wattless startup and shutdown process,the prediction accuracy of the system dynamic model for the operating conditions with power,the temperature dynamic response speed of the large time-delay system,and the control accuracy of oxygen excess ratio under the condition of parameter uncertainty.(2)A combined dynamic modeling method consisting of a hydrogen flow resistance network model and an equivalent circuit model is proposed to evaluate the degradation rate distribution in the multi-cell stack during startup.Experimental results show that the combined model can reflect real physical phenomena,explain the coupling mechanism between the distribution of reactive gas and the distribution of current and voltage in the multi-cell stack,and characterize the internal relationship between the capacitance effect and the dynamic change of voltage.A startup and shutdown strategy is proposed from two aspects of auxiliary load and gas purging sequence based on the simulation results to reduce the local high potential caused by the hydrogen/air interface and capacitance effect,the time for the average single cell voltage to decrease to 0.4 V during the shutdown is shortened by about 500 s.(3)A high-precision dynamic model of the fuel cell system including the fuel cell stack,the gas supply systems,and the heat management system is developed by the hybrid modeling method.For the fuel cell stack,three output characteristics models with different parameters are developed with limited experimental data,the model parameters are optimized based on the harmony search algorithm and four improved particle swarm optimization algorithms.A machine learning modeling method based on different working conditions is proposed by establishing the mapping relationship between operating parameters and model parameters,which reduces the generalization error of the output characteristics model by 58.3%.The simulation and experimental results show that the system model can accurately describe the dynamic characteristics of the system and be applied to the study of control strategies.(4)A control strategy containing dual-stage actuators is proposed to address the problem of slow temperature response of the large time-delay system.A fuzzy PID controller is designed for heat dissipation power regulation,and the different evolution algorithm is applied to search for the proportion factor that minimizes the tracking error to improve the stability of the temperature control system.ADRC strategy for coolant flow regulation is designed and the key parameters of the controller are calculated.The disturbance estimation and compensation capabilities of the algorithm contribute to improving the temperature response speed under load disturbance.Tracking differentiators are introduced in the forward channel and feedback channel to reduce the influence of temperature measurement noise.The research results indicate that the proposed control strategy reduces the temperature regulation time by more than 23.8 s.(5)A MRAC-based oxygen excess ratio control strategy is proposed fully considering practical issues such as the time-varying of parameters and load disturbance.Reference models based on PI and LQR algorithms are designed and compared,and the impact of the adaptive gain on control performance is analyzed.An MRAC control strategy including proportional and derivative elements is proposed where the proportional element compensates for the shortcomings of traditional MRAC in dynamic response speed,and the derivative element effectively suppresses load disturbance and pressure overshoot.The research results show that the proposed strategy has better robustness,and can make the operating trajectories of the air compressor far away from the surge boundary under different parameters,the wear of the air compressor is reduced by 37.7% compared to traditional MRAC.