Research On Performance Evolution Evaluation,health Management And Control Of SOFC

Due to the complex structure of the Solid Oxide Fuel Cell system(SOFCs),its performance is difficult to be guaranteed.Performance evolution and faults may occur in various subsystems and components,and many faults are gradually evolved by small defects(e.g.serious gas leakage,failures caused by increased resistance,etc.).In this case,how to ensure the efficient,stable,and long-life operation has become one of the bottlenecks that need to be resolved in the commercialization of the SOFC system.Therefore,the research aims to realize accurate,reliable,quantitative performance evaluation considering the performance evolution of SOFC,while putting forward health management and control strategy to guide actual SOFC system design and control development.The research contents are as follows:The gas leakage of SOFC is a gradual evolution process.Slight gas leakage may lead to insufficient reactant supply and greatly accelerate the aging behavior.Serious gas leakage may lead to heating out of control.Therefore,a model-based real-time gas leakage diagnosis scheme is proposed for the common leakage faults that occur at the stack outlets on the k Wlevel SOFC test platform.According to the existing stack of the Fuel Cell Research Center in HUST,based on the balances of thermal,electric,and pneumatic,a multi-physical dynamic mechanism model of SOFC stack is established,and the model parameters are calibrated through the experimental data.Based on the stack model,the fuel and air leakage signals are reconstructed online by an adaptive covariance matched Unscented Kalman Filter(UKF).At the same time,an adaptive threshold generator is proposed to generate dynamic thresholds to enhance the robustness of gas leakage fault isolation.The simulation and experimental results show the effectiveness and practicability of the proposed gas leakage diagnosis scheme.Complex degradation mechanisms,such as coarsening,oxidation,and poisoning,lead to the evolution of the SOFC performance,which has become the main obstacle to the longlife operation of the SOFC system.Therefore,aiming at the problem of how to accurately evaluate the performance degradation,the feasibility of using degradation resistance to evaluate the health state of SOFC is proposed and verified.In addition,in order to eliminate the interference caused by directly using the measurement and electrical characteristic model to calculate the health state,a real-time estimation method of the SOFC health state based on UKF is further proposed.The online estimation results of degradation resistance using UKF algorithm under constant and system conditions prove the feasibility and effectiveness of the health state evaluation method.Due to the degradation evolution of SOFC health state,the failure will occur when the health state evolution develops to a certain extent.In this case,the prediction of End of Life(EOL)and Remaining Useful Life(RUL)play a positive role in improving the reliability and durability of SOFC.Therefore,considering the prediction problem of SOFC health state degradation evolution,the SOFC degradation process model is developed and verified from the perspective of mechanism analysis.On the basis of this model,a SOFC fault prognosis method implemented by particle filter is proposed.In the case study,this method effectively predicts the health state evolution trend of SOFC and then gives reasonable EOL and RUL prediction values.As the problem of SOFC performance degradation cannot be ignored,it is necessary to avoid performance degradation in the stage of system design and control development.Therefore,according to the power demand of the monitoring station,the SOFC health management and control method oriented to long-life applications is proposed.At the hardware level,the SOFC system and the battery are indirectly connected in parallel through a unidirectional DC-DC converter;At the software level,the collaborative control and management methods among thermal-electric-pneumatic are proposed to reasonably control the energy flow between SOFC and battery while avoiding excessive temperature and fuel starvation,so as to prevent the performance degradation and prolong the service life.The simulation results show that the proposed hybrid system can meet the power demand of the monitoring station,and the developed thermal-electric-pneumatic co-controller can not only reasonably distribute the energy flow but also effectively ensure the system’s long-life operation.