Study And Realization On Low Speed AC Motor Servo System

Nowadays, low speed and sometimes even ultralow speed electric machine drive servo systems with exquisite steady and dynamic performance have been highly demanded for wide range of particular applications, such as modern high precision computer numeric control (CNC) machine tools, automatic tracking systems of radar and satellite antenna, industrial robotic systems, and optical processing equipments. The electric machine in those servo systems which require good tracking and anti-disturbance performance in low speed operation should be able to operate smoothly at certain speed or load changes. Normally, the performances of electric machine drive servo systems are inevitably deteriorated by various quantization, sampling, and measurement errors as well as machine torque ripple. Furthermore, the precision and dynamic response of the speed feedback in such servo system might be greatly tampered in low speed operation so that high-performance speed control of the machine could be rather difficult to accomplish. Consequently, the low speed performance of such servo system could be greatly compromised as the machine will usually spin and stop randomly. Overall, the smoothness of low speed electric machine drive servo system is determined by two factors:torque characteristics of the machine and performance of the controller. Investigations on electric machine itself, driving scheme, control strategy, and sampling method of position sensor are usually indispensible in order to improve both the steady and dynamic performances of the low speed electric machine drive servo system.Firstly, the paper has covered background and corresponding technical fields of the low-speed electric machine drive servo system. An overview of the state of the art in this research area has been also presented and meanwhile both the context and significance of such topic have been declared.Secondly, the experimental platform of low speed electric machine drive servo system has been built and introduced. A slotless permanent magnet synchronous machine is employed in this system in order to eliminate the cogging torque and minimize electromagnetic torque ripple. Comprehensive analysis on the structure and electromagnetic field has been carried out to unveil the unique features of the slotless machine so that mathematic model of the machine can be established based on rotor magnetic field. Stator voltage vector control with predictive current method has been employed to alleviate current distortion and hence torque ripple.Thirdly, a first-order disturbance observer based on filtered position signal has been presented and implemented in the proposed experimental platform. A first-order low pass filter, which filters the output value of the disturbance observer instead of stator winding current, is employed to achieve pseudo-differential operation so that the computational speed error caused by position quantization error of the optical encoder can be reduced over the low speed operation region. Besides cogging torque free, the slot less permanent magnet synchronous machine has a relatively larger air gap and hence small inductance and electrical time constant, resulting in excellent dynamic performance. Consequently, the low speed slotless permanent magnet synchronous machine drive servo system with proposed disturbance torque observer can effectively suppress torque disturbances and improve the drive performance. A stable low speed region (around lr/min) has been achieved in the proposed servo system equipped with a moderate resolution (1000ppr) encoder. Both the simulation and experimental results have revealed that the proposed disturbance torque observer can be easily implemented in practical applications to suppress the corresponding effects of the disturbance torque and computational speed error.Fourthly, a speed calculation method based on second-order sliding mode differentiator has been proposed. The method imposes twice sliding mode operations on the position sampling signal to retrieve the rotor acceleration information. Thus, the rotor speed can be accordingly obtained accurately with on compromise on the effectiveness and timeliness in low speed region. The dynamic performance of the servo system in low speed region (20r/min-40r/min) can be effectively improved by the proposed second-order sliding mode differentiator together with the current compensation algorithm of disturbance torque. The experimental results have shown that the proposed method cannot only improve the dynamic response and robustness of the servo system against the load disturbance.The most direct and effective approach to improve low speed performance of a servo system is to increase the resolution of the encoder. Last, a novel interpolation method for sinusoidal quadrature encoder, which involves approximate-linear function and interval division algorithm, has been introduced. The sine and cosine wave signals received by the servo system can be converted to high-resolution pulse sequence by this method, so that the speed of the machine can be accordingly calculated. The method can be easily implemented in the system to achieve high precision and anti-interference in speed calculation with little extra computational resource but no additional hardware expense.