Research On Torque Ripple Suppression Of Brushless DC Motor Based On Iterative Learning

Brushless DC motor is a fusion of modern electronic technology and modern control theory.It is a mechatronic product that comes with semiconductor technology.It has aroused widespread concern because of its good speed regulation,high power density,good reliability and simple control.However,its inherent periodic torque ripple problem is a limitation of its application in some high precision requirements.The torque and speed ripple suppression of the motor under low speed loading are the main research in this thesis.A controller based on iterative learning algorithm is designed for this problem and satisfactory results are obtained.Initially,the principle of the motor in the two-phase conduction mode is introduced.The general mathematical expression of brushless DC motor and mathematical model have been established,and the transformation formula of three-phase state model to two-phase stationary coordinate system model have been given based on the circuit topology of three-phase full-control bridge-type star connection.What’s more,in view of the problem of periodic pulsation of the torque and speed of the motor at low speed,the direct torque control model is applied.The mathematical principle and modeling process of voltage and current conversion module,stator flux linkage calculation module,sector division and selection module,and space voltage vector selection module are introduced in detail.The real-time torque of the motor can be obtained,and the control of the motor torque and the suppression of the current can be realized by using the phase-phase current to calculate the electromagnetic torque of the motor.Due to the direct torque control can not solve the problem of periodic torque ripple of brushless DC motor very well,and the iterative learning algorithm is very suitable for such repetitive,periodic and uncertain system,so the control method of iterative learning is chosen.According to the control idea of iterative learning,a PI type iterative learning controller is designed,which is parallel to the speed controller to form a composite controller,so that the given value of the torque can be compensated in real time.Such a controller not only can effectively suppress the commutation torque ripple,but also has a certain suppression effect on other types of torque ripple,and is also helpful for the fluctuation of the rotation speed and the tracking accuracy.Finally,experiments were carried out using the experimental system platform of micro control system STM8S series in order to further verify the feasibility and practicability of the iterative learning algorithm.Using the control method of the experimental platform and Matlab software,the experimental comparison model between the iterative learning system and the classic PID is established in Matlab.The communication between the upper and lower machines uses the serial port module.According to the simulation calculation,the motor control quantity is obtained.The data is transmitted to the control chip STM8S through the serial communication module.The motor is runed by the control signal,and the actual speed value of the motor is collected and transmitted back to the model through the serial port to realize the waveform comparison between the set value and the measured value.It is concluded that the iterative learning controller has a good effect on this periodic speed torque ripple reduction,the torque error is reduced by 90%,and there is no speed ripple,by analyzing the simulation waveform.It can be seen from the experimental results that compared with the classical PID control,the iterative learning controller has higher speed tracking accuracy under steady state,and can realize precise tracking of the rotation speed and no pulsation of the rotation speed,and the motor runs more smoothly.