Research On Key Technologies Of Permanent Magnet Sysnchronous Motor Engineering Servo System

Permanent magnet synchronous motor(PMSM) has many advantages, such as small torque pulsation, high torque/inertia ratio and high efficiency, which make it hold the pivotal status in the modern alternating current motor. Along with the development of power electronic technology, control technology, the motor technology and so on, PMSM driving system is developing toward high power engineering direction. The servo system applied in this high power engineering is named as PMSM engineering servo system in this paper. It has some special key technologies, such as heavy load drive, high power, anti-disturbances, sensorless control and so on. This article makes some research and exploration on three key technologies of all, that is heavy load driving technology, anti-disturbance adptive control technology and sensorless control technology.The heavy load technology is the core technology to raising the flexibility level of traditional mechanical press. Research on the technology of PMSM engineering servo system is expressed taking the crank servo press as the object. Firstly, the design scheme of crank servo press prototype and dynamic system drived with heavy load is introduced respectively in this article, with the combination of PMSM and capacitor energy storage device substitutes for the traditional mechanical press's combination of asynchronous motor and mechanical flywheel– coupling. The load characteristic of press is analyzed based on it. In order to let the designed PMSM drive system meet the dynamic request of press driving by heavy load, theoretical analysis on parameters of the dynamic system, that is PMSM electrical parameters and capacitance value of the capacitor group about the drive system, is carried out separately. The parameters'approximate calculation plan and analytic expression are also given. In order to save the development cost and cycle, analysis on the scheme and feasibility of the system which has not be established yet is carried out. A simulation software of crank servo press electro-mechanical integration is developed combining C++Buliler with Matlab. Taking an 80 ton crank servo press as the application object, the rule between capacitor energy storage parameters and motor operating velocity is explored, and the design scheme of capacitor energy storage parameters is proposed which is suitable for servo press's part controllable operating. Analysis is done based on the constant speed driving mode of servo press, and a design scheme of capacitor energy storage parameters selection and a group of reasonable parameters are proposed. Based on the parameters, to realize variable speed driving and multi-adaptability of the press, it puts emphasis on detailed analysis of the servo driver's control parameters and the press's output ability. And important conclusion is also provided. Design the hardware device of the driving system (Experiment Platform 1) and simulation and experimental confirmation are carried out on it. The results prove that the approximate analytic formula and design scheme of dynamic system parameters are reasonable; the simulation software developed is correct; the conclusion obtained from the simulation analysis based on the servo motor driving at constant speed and variable speeds is also correct. According to the above, a set of capacitor energy storage hardware device is developed as the energy storage device for press ramming processing. It has not only realized the accelerative function of traditional mechanical press's mechanical flywheel, but also slowed down the impact of the summit current on power grid during the process of ramming.The PMSM drive system based on the energy storage device can satisfy the power and control request of servo press driving with heavy load. It lays the foundation for furthermore realizing the optimization of dynamic system parameters and raising the flexibility level of mechanical press.The performance of PMSM engineering servo system is influenced by the load characteristic parameters'changes( that is to say disturbances), due to the high power and heavy load and bad condition of operating site. Taking suppressing the disturbances from load torque and inertia as the goal, an adaptive control method with anti-disturbance function is proposed based on the thought of identifying some disturbance and then compensating or auto-tuning to suppress the disturbance. And the foundational question is the effective observation of disturbance. In view of time-variable inertia, two identification methods based on Landau idea and Gradient idea are designed; in view of time-variable load, a simple reduce-order load observer is proposed to identify the load torque. Each simulation model is constructed and the control platform for AC motor based on Freescale 56F8346 DSP is designed (Experiment Platform 2). Testing on a surface mounted PMSM, the results show the accuracy of respective algorithms. The convergence time of Landau algorithm is better than Gradient algorithm, but the identification precision of Gradient algorithm is higher than Landau algorithm; the load torque observer can identify the magnitude of load torque effectively, and the dynamic response performance is better. The adaptive control method uses Landau algorithm to identify inertia, regulates the controller parameters by estimated values, and compensates the results obtained from load observer feedforward the AC current side simultaneously. From the experimental results we can see, the parameters'self-regulating method based on Landau algorithm can refresh the parameters of the controller in time while the inertia changing. Comparing to the system without parameters self-regulating, the speed response performance is improved. Compensating the velocity disturbance caused by changing load improves the response performance of velocity by estimatied load torque value. Therefore, the adaptive control method proposed in this paper has the inhibitory action on the inertia and load torque disturbances, and can apply in some drive occasions with time-variable loads and inertia changing slowly.In order to solve the limitation problem occurred on physical sensor in the operating occasions with bad condition, research on the key technology of PMSM sensorless control, that is to say, rotor position estimation is expanded first. The rotor position estimation method, that is to say, sliding observer is studied which suited to middle and high speed situations. It is aim at expanding the domain of low speed operating. According to the references looked up presently, in view of the existed questions during the improvement of some methods based on traditional mathematical model sliding mode observer, a novel improved sliding mode observer algorithm based on the mathematical model is proposed. The back-EMF estimated value is fed back to the stator current's observation calculation. It can improve estimate precision of rotor position at low speed and improve system stability at middle and high speed through choosing reasonable feedback value. The system simulation model is constructed in the article, and taking a surface mounted PMSM as the object, analysis is done on the Experiment Platform 2. The results show that the minimum speed of traditional sliding observer realizing the rotor position estimation is 50r/min, and under the same condition the low speed of mathematical model improved version sliding mode observer may expand to 8r/min; when operating at middle or high speed, the estimation precision and stability is better, and has the very strong robustness to the variation of load. Rotor estimation method suited to zero-speed occasions, that is to say, voltage pulse vector method is researched. Its detective principle and implemental strategy are analyzed detailed. The controller (Experiment Platform 3) is designed and experiments are carried out on it. The results confirm the validity of the algorithm.Secondly, under the above research foundation based on the mathematical model improved version sliding mode observer estimating rotor position, the adaptive control method for rotor position eastimation while motor operating at wide speed range is proposed, to realize the PMSM sensorless closed loop vector control in a real sense, which is also named as the adaptive principle of the back-EMF feedback gain coefficient. Phase lock loop is used to control rotor speed. For the confirmation of algorithm's validity, tests are carried out on the Experiment Platform 2. The experimental result show that, mathematical model improved version sliding observer based on the auto-adapted principle and phase lock loop control can estimate the rotor position and speed accurately while PMSM operating at the low speed (5r/min); sensorless sensor vector control velocity modulation system based on this algorithm has the better dynamic response and estimated rotor position and actual rotor position coincidence while changing speed operating at the middle and high speed domain; in the low speed domain, estimated rotor position can track the actual rotor position effectively, and in the range above 15r/min(1 percent of rated speed) the sensorless sensor control can replace the sensor control. This sensorless sensor vector control system has the characteristic of wide speed range and strong robustness.