Research On Humidification Control Technology Of Kilowatt Fuel Cell Test System

The fuel cell in the field of hydrogen energy has broad development prospects in the field of transportation,because the reaction process is not restricted by the Carnot cycle.While the fuel cell test system plays an important role in promoting the industrial application of fuel cell,its humidification has multiple problems such as large hysteresis and non-linear characteristics,and difficulties to meet the system requirement on the dynamic settling time.This thesis focuses on the humidification control technology of kilowatt fuel cell test system.The main research contents and results are as follows.A research platform for the characterization of bubble humidifiers was designed and built.The hardware of the platform was designed,the structure of the bubble humidification tower was introduced,and the hardware construction of the platform was completed according to the requirements and operating conditions of the humidifier in the test system.Then,programs of the controller and the host computer were designed,and the software development of the platform was completed.The input and output characteristics of the bubble humidification system were analyzed,and a semi-mechanical simulation model of the system was built.According to the experimental data of the platform,the influence of the system output characteristics on the gas flow,pressure,temperature and liquid level of deionized water was studied.According to the influencing factors of output characteristics,the humidification simulation model was divided into three parts: fluid dynamic behavior model,water molecule mass transfer model and convective heat transfer model,and the structure of the model was analyzed.Based on the structure of the simulation model,the unknown parameters in the simulation model were identified by the experimental data,and the simulation models of saturated humidification and unsaturated humidification were established.The simulation results of the two simulation models were compared with the experiment,and the correlations between the simulation data and the experimental data were 0.978 and 0.989.The simulation model of humidification system was validated.The mathematical model of bubble humidification was established,and the humidification control strategy was designed and implemented,which accelerated the dynamic adjustment of humidity and met the requirements of the test system.The simulation model of saturated humidification was converted into a state space model;the state space model of unsaturated humidification was identified by the N4 SID method,and the Wiener state space model of unsaturated humidification was obtained by using particle swarm algorithm.Based on the state space model and the Wiener state space model,the PID control strategy and model predictive control strategy of saturated humidification and unsaturated humidification were designed and implemented.The dynamic performance,steady-state performance and anti-interference ability of the control strategy were evaluated through simulation experiments.The results showed that the settling time of adaptive model predictive control in saturated humidification and nonlinear model predictive control in unsaturated humidification were 90 s and 69.6s,which was better than PID control obviously.