Optimized Starting Control And Low Speed Torque Ripple Suppression Of Brushless Servo Motor

Due to its simple structure,reliable operation,small size,high efficiency,low loss advantages,coupled with the development of power electronic devices and frequency conversion technology,brushless servo motor is widely used in aerospace,intelligent electrical appliances,electric vehicles,ships and automatic factory production line servo system.The detection accuracy of the initial rotor position will affect the loading and starting characteristics of the motor,which is one of the important conditions for the smooth start of the motor,and is also the hot spot and difficulty in the current international research.If the deviation of the initial rotor position detection occurs,it will directly lead to the weakening of the load capacity of the motor starting,starting failure or reverse phenomenon.The stable output of motor torque is the control strategy and also the key element to ensure the stable operation of the system.This paper first introduces the brushless servo motor in the modern application field and development status,introduces the mainstream control mode,and focuses on the starting optimization control and low speed torque ripple suppression algorithm proposed by domestic and foreign scholars,according to the structure of brushless servo motor,the brushless servo motor drive model is established.The double closed-loop vector control algorithm and the core motor control strategies such as coordinate transformation and SVPWM modulation are selected.On the optimization of drive algorithm,the solutions of speed and torque pulsation in start-up stage and low speed running stage of brushless servo motor are proposed respectively.In terms of the optimization of starting stage control,a starting scheme combining high frequency pulse vibration signal injection with position observer is proposed.Compared with the traditional open-loop IF starting mode,the starting performance is optimized and the possible risks of oscillation and reversal are avoided in the starting process.On the low speed torque ripple suppression,two kinds of torque suppression methods are proposed,namely Romberg reduced order load observer and reduced order extended Kalman filter algorithm.Rompberg reduced order load observer combined with feedforward control introduces load torque into the current loop,so that the current regulator can change with the load in real time.Since the electromagnetic torque is proportional to the current,the load torque can be adjusted with the movement.The reduced order extended Kalman filter optimizes the structure of traditional Kalman filter algorithm,simplifies the calculation process,improves the calculation speed,and extends the application scope to nonlinear domain.Finally,an adaptive non-singular terminal observer and a sliding mode observer combined with radial basis neural network are proposed to control the brushless servo motor from low speed to high speed.The terminal sliding mode has the advantage of finite-time convergence.The non-singular terminal sliding mode optimizes the structure of the traditional sliding mode observer,and the selected switching function has no singular value in the domain,which avoids the oscillation.The RBF neural network can provide gain coefficient for the adaptive terminal sliding mode observer rapidly and improve the fast response and robustness of the system.The experimental platform based on TMS320F28069 DSP is adopted in the experiment,and the results test the effectiveness of the improved algorithm mentioned above.In different working conditions,various algorithms were used to carry out variable frequency speed control experiments,and the speed and torque following performance of each algorithm was compared and analyzed,which verified the correctness and effectiveness of the proposed control algorithm.