Study On The PEMFC Neural Network Model And Fractional Order PI~?D~? Control Based On An Improved Krill Herd Algorithm

The problem of energy shortage and environmental pollution have become two major focus problems in the world.Proton exchange membrane fuel cell(PEMFC)has become a very promising clean power supply device due to its advantages of cleanness,high efficiency,sustainability,low noise,low temperature and fast start-up.This paper combines Elman neural network and improved krill In this paper,the air-cooled PEMFC system is taken as the research object to study the effective control of the reactor voltage.The main contents of this paper are as follows:1. Under certain assumptions,the voltage model of PEMFC is constructed based on the electrochemical theory combined with the double layer capacitance effect,the cathode gas partial pressure model and the anode gas partial pressure model;the voltage current characteristics of PEMFC are analyzed based on the theoretical formula and the voltage model;the influencing factors and control variables of the voltage response of PEMFC under the condition of changing load are determined to model and design the voltage control system for PEMFC It provides a certain theoretical basis.2. The voltage model of PEMFC is established by Elman dynamic neural network. For the problems that the hidden layer structure of the model is difficult to be determined and the initial parameters may be generated randomly,the model is easy to fall into local extremum and the results cannot be reproduced,this paper proposes an ikh algorithm to optimize the structure and initial parameters of Elman neural network.The simulation results show that the optimized Elman neural network model has the advantages of small network scale,high fitting accuracy and fast convergence speed.This modeling method is also suitable for water-cooled PEMFC modeling and has certain universality.3. On the basis of krill herd(KH)algorithm,an IKH algorithm is proposed by introducing the cosine decrement step and the selection operation of simulated annealing mechanism.Through the simulation test of six benchmark functions,it is verified that the optimization algorithm has better optimization accuracy and faster convergence speed.4. Considering that the PID controller is widely used in the practical industrial field,the fractional order PI~?D~?controller has the same structure and more flexible control effect.In order to solve the problem that the parameters of fractional order PI~?D~?controller are too many to be adjusted,this paper uses the IKH algorithm to optimize the control parameters of fractional order PI~?D~?controller.The integral order PID control parameters and fractional order PI~?D~?control parameters are optimized by ikh algorithm.Simulation tests show that the IKH algorithm optimized fractional order PI~?D~?controller can control the hydrogen flow more quickly when the load of PEMFC changes,so that the output voltage of the stack can reach and maintain at the expected voltage value quickly and stably.