Modeling And Optimization Of Fuel Cell Cold Start System In Low Temperature Environment

In the context of serious environmental pollution and increasingly depleted fossil energy,proton exchange membrane fuel cells are considered to be one of the most potential power sources for future vehicles due to their advantages such as clean products,high energy density and short energy recharge time.One of the major technical challenges in the commercialization of proton exchange membrane fuel cells is the ability to successfully start in sub-zero environments,often referred to as the "cold start" problem.The main problem of the cold startup of proton exchange membrane fuel cells is that the water generated during the startup process will freeze below zero.If the catalytic layer is completely covered by ice before the reactor temperature rises to 0 ℃,the cold startup will fail and irreversible damage will be caused to the proton membrane.This paper takes proton exchange membrane fuel cells as the research object to study the cold start system modeling and controller design.Firstly,in order to describe the temperature change,the ice volume fraction change process and the reactor operating state during the cold start-up of proton exchange membrane fuel cells,a simulation model of the cold start-up system of fuel cells is established based on the modular modeling idea,including the temperature rise model,water content model and output voltage model.On this basis,the effects of ambient temperature,reactor current and anode oxygen flow rate on the cold start effect of the cold start system are analyzed.Through a reasonable simplification of the cold start system mechanism,a control-oriented three-order system model is established and verified.Then,aiming at the problem that the ice volume fraction in the cold start system of fuel cells cannot be measured,an estimation method of ice volume fraction based on the extended state observer is proposed,taking the reactor temperature as the correction term.Aiming at the uncertainties of unmodeled dynamics,internal and external perturbations and parameter perturbations in cold start systems,a perturbation observer based on radial basis function neural network is designed to estimate and compensate the uncertainty information online,and the stability of the observer error system is proved under the Lyapunov stability framework.The simulation results verify the effectiveness of the state observer and disturbance observer.Finally,according to the characteristics of multi-input/output,strong nonlinearity and multiple constraints of fuel cell cold start system,a cold start time optimization control method based on nonlinear model predictive control is proposed.Combined with the influence of reactor current and anode oxygen flow on icing amount and heating rate in cold start system,an adaptive weight coefficient setting method related to ambient temperature is proposed.In order to further improve the rapidity of cold start system,a global optimization strategy for the weight coefficient of controller based on inverse distance weighting coefficient and radial basis function is proposed.The cold start simulation experiments with different controller weight coefficients are carried out to verify that the proposed method can effectively shorten the cold start time of the fuel cell system.