Research On Neural Network Control Of Controllable Excitation Linear Magnetic Suspension Synchronous Motor

The background of the project is the application of the magnetic suspension feed platform.Under the support of the National Natural Science Foundation project “Study on the Operation Mechanism and Control Strategy of the Controllable Excitation Linear Synchronous Motor Maglev Feeding Platform(Project Approval Number: 51575363)”,the main research is on the controllable excitation line.Magnetic suspension synchronous motor neural network control.The main research work of this paper includes the following aspects:1.The establishment of a mathematical model of a controllable excitation linear magnetic levitation synchronous motor.Through the in-depth analysis of the working principle and structure of this kind of motor,the electromagnetic force equation and the motion equation are derived.Combined with the equation of motion,voltage equation and flux linkage equation of the motor in the shaft coordinate,the mathematical model of the feed system is derived.2.Research on Neural Network PID Control of Controllable Excitation Linear Magnetic Levitation Synchronous Motor.Introduce the principles of neural networks,the structure of biological neurons,mathematical models of artificial neurons,excitation functions,and learning rules.Combining the neural network control method with the traditional PID control,the BP neural network is used to adjust the three parameters of the output,and the system is controlled by the self-learning of the neural network and the adjustment of the weight.And use Matlab to simulate,get the input step signal,sinusoidal signal,and the speed response curve in case of sudden addition.3.Research on Decoupling Control of PID Neuron Network for Controllable Excitation Linear Magnetic Levitation Synchronous Motor.This paper introduces the characteristics,structure and calculation methods of PID neural network.The structure and algorithm of multi-output neural network are studied and analyzed.A two-output PID neural network structure is designed for the decoupling problem of controllable excitation linear maglev motor.The simulation is performed by Matlab to obtain the decoupled horizontal velocity and The output curve of the levitation height.