Control method of disc type switched reluctance motor

Abstract

The invention discloses a control method of a disc type switched reluctance motor. The control method comprises the steps of establishing a nonlinear torque control model, establishing a multivariable static neural network controller model, establishing a feedback control system and performing neural network driving control. According to the method, the actually measured data of the axial magnetic field disc type switched reluctance motor is analyzed, the influence of the rotating speed, the current, the turn-on angle and the turn-off angle of the motor on the output torque is described, it is considered that the turn-on angle and the turn-off angle need to be dynamically adjusted according to parameters such as the current, the rotating speed and the torque, and a static multivariable neural network controller model is established; a training method of a neural network controller is designed, and a variable step size optimization method based on a least square method is provided for quickly acquiring optimal neural network online training data, so that an axial magnetic field disc type switched reluctance motor control system based on current optimization is realized, and the neural network controller is combined with a PID regulator, so that the whole system has dynamic adjustment capability.

Description

technical field [0001] The invention relates to a control method of a disc-type switched reluctance motor. Background technique [0002] Ordinary radial field switched reluctance motors have the following characteristics: the stator and rotor are double-salient pole structures, and there is no winding or permanent magnet on the rotor. The windings work independently, can operate in a lack of phase, and have strong fault tolerance; the switching elements of the power converter are connected in series with the windings, so there is no danger of direct short circuit between the upper and lower bridge arms, the drive is simple, and the reliability is high; the starting current is small and the starting torque is large, suitable for Occasions that require frequent starting; many controllable parameters, flexible control, good speed regulation performance, wide speed regulation range and high efficiency range. [0003] Combine the switched reluctance motor with the axial magnetic...

AI Technical Summary

Problems solved by technology

It is worth noting that the current neural network PID controller generally achieves the purpose of controlling the output of the motor by adjusting a certain parameter (the input of the motor), but the switched reluctance motor has many controllable parameters (turn-on angle, turn-off angle, Current, etc.), the output speed and torque will in turn affect the setting of the above parameters, so the current neural network PID controller has certain limitations in the application of switched reluctance motor control

Images