Output Optimization Control Of Pemfc Based On Online Identification Model

Proton exchange membrane fuel cell has become one of the new energy sources for people to deal with traditional energy problems due to its high efficiency,cleanliness,and good output stability.It has been widely used in various fields,such as rail transit,military industry,aerospace and so on.Open-cathode PEMFC,as a representative of the simple structure of the system,has become the most rapidly developing fuel cell due to its high power density and rapid start-up at low temperature.At present,there are a large number of simulation modeling results on fuel cell research.Since the open-cathode PEMFC is a complex dynamic system with multiple inputs and multiple outputs nonlinearly coupled,the controller scheme obtained through pure theoretical simulation is difficult to be used in he hardware system and the simulation modeling does not have strong practical application capabilities.In the actual system platform,the operational data of the actual measurement system is analyzed and processed to obtain the output power characteristics of the system based on the actual operating of the open-cathode proton exchange membrane fuel cell.Based on the output power characteristics of the system and the system structure characteristics of the open-cathode PEMFC,the test control platform is designed.The hardware measurement and control platform composed of the control board and the LabVIEW PC software is built.Through the design of the system start-stop control scheme and system flow,the stable and efficient operation of the stack is ensured and the operating characteristics of the open-cathode PEMFC system are analyzed and verified on the built monitoring and control platform.The system optimal output power regulator is designed to obtain the optimal output characteristics of the system.The performance optimization control method of PEMFC system based on the optimal output power characteristics of the system is proposed.Discrete incremental PID,fuzzy logic control and adaptive fuzzy PID control combining the advantages of both are used to realize the optimal output power control.Through experiment,the influence of each control parameter on the system control effect is analyzed,and the reasonable control parameters are selected to realize the optimal output of the system.Comparing the dynamic response characteristics of different control methods under load changes,the conclusion that adaptive fuzzy PID control has the best system dynamic response and steady state characteristics is obtained.An output performance optimization control method based on online identification is proposed,considering the nonlinear time-delay characteristics of the system.The dynamic linearization of the nonlinear system is used to obtain the general model of the system.The model-free adaptive predictive control(PFDL-MFAPC)dynamic linearization based on partial format is adopted combined with the idea of predictive control.Then completes the theoretical derivation and proof of linearization method.Based on the designed optimal criterion,the online identification of pseudo-gradient vector parameters in the general model is realized,and verify the effectiveness of the algorithm through the measurement and control platform.Compared with the adaptive fuzzy PID control method,the proposed algorithm can effectively reduce the system overshoot,make the output power of the system more stable,and the fluctuation of the fan control voltage is smaller,which is more conducive to the long-term stable operation of the stack.