Parameter Identification Algorithm And Verification Analysis Of Hybrid Excitation Synchronous Motor Based On ANN

The hybrid excitation synchronous motor takes into account both efficiency and performance,and has a broad application prospect.However,in the actual application of the machine,because of many elements for instance corrosion,radiation,motor aging,load changes,etc.,the machine parameters can be different from before as a result.This will have influence on the running status of the system.For the sake of mastering the changes of parameters in current time and convenient control of the system in current time,we need to tell the difference of the motor parameter in the system.Based on the vector control system of HESM,this paper researches how to recognize the parameters online.First,this paper sets up the HESM’s model in each coordinate axis,and derives each equation.How to control the motor is explained and selected,and use Simulink and System Generator to complete the HESM system modeling and simulation respectively.The simulation results of electrical angle,excitation current and so on were got,and the simulation waveform was correct and the result was obtained.Better,it proved the validity of the HESM vector control system,and established the base for the following chapters.Secondly,it recommends the BP neural network used as a way of how to tell the difference of parameter,and elaborates its way of modeling and parameter training in detail.On this basis,System Generator is used to model and simulate the how to tell the difference of parameters online based on BP-NN.The error of result is almost zero.Third,after the simulation verification of the HESM vector control system model and telling the difference of parameters model on account of BP-NN,the two are combined to complete the online telling the difference of parameters of HESM according to BP-NN.The System Generator model was built,and the magnitude of error which was got by simulating it is between 10^-2 and 10^-7,which meets the accuracy requirements.Fourth,based on the chip,XILINX’s Artix-7 chip XC7A200TFBG484-2,a hardware verificated condition is completed and the test platform could be set up.Then,various hardware modules were designed and produced,and the System Generator tool was used to develop the system and generate the corresponding IP.And writing the initialization code of each chip and the overall control code.On this hardware platform,the validity of the parameter identification algorithm of HESM based on BP-NN is proved.The validity and authenticity of this article is verified by the experiment,which lays a foundation for its engineering application.