Research On Nonlinear Control Strategies And Sensorless Methods Of PMSM

Permanent magnet synchronous motors (PMSMs) have been widely used due to its rugged construction, easy maintenance, high efficiency, and high torque to current ratio, low inertia. However, the dynamic mathematical model of the PMSM is a nonlinear, time varying and multiple variables object, and the control system is often influenced by uncertainties which are induced by parameter variations, external load disturbances, unmodelled plants and nonlinear dynamics. Although the vector control has Been applied to PMSM, the conventional control methods have some disadvantages. Besides, the control of PMSM needs a speed sensor to calculate and obtain the speed and position, but this may increase the complexity, weight and cost of the system. The research and development work on the sensorless driver is progressing greatly. This dissertation relies on the Research of Energy Saving Technology of Electrical Driven and Control System (Project NO. is2006AA04Z183), which is supported by National High Technology Development Plan. In this dissertation, some nonlinear control strategies and sensorless methods of vector controlled PMSM system is proposed and discussed. The main research work is as follows:1. A backstepping controller with on-line adaptive load torque observer is designed forthe speed control of PMSM, which has great robustness to load torque variation; furthermore, a backstpping controller based on sliding mode variable structure is proposed and designed for the speed and position control of PMSM. Sliding mode variable structure with exponential reaching law enhances the quick response and anti-disturbance ability of the control system. At the same time, extended state observer is designed to evaluate the load torque of the system online, so the control process can adjusted timely, which lead to the parameters of exponential reaching law being smaller.2. The mathematical model of PMSM is decoupled into a second-order linear speed subsystem and a first-order linear current subsystem based on input-output linearization theory, then a controller is designed for the speed and current control of the PMSM, which makes the tracking performance of given speed and current is good. Besides, an observer based on MRAS is designed to observe the variation of stator resistance and compensate the controller, so the control performance is improved. Furthermore, a method combined with sliding mode variable strucuture and feedback linearization is proposed and designed which has better tracking performance for the given speed and current. As the load torque is used for the derivative of speed and the controller, then it is observed based on extended state oberver.3. An adaptive dissipative Hamiltonian controller with load toque observer is designed from the view of energy balance for the speed regulation of PMSM. When the motor is running smoothly at the given speed, the minimum value of Hamiltonian function is obtained, then we could get the desired speed regulated performance. Besides, a dissipative Hamiltonian controller with a Luenberger observer for flux linkage and load toque is proposed and designed, and the robustness of the control system is enhanced.4. Based on active disturbance rejection control (ADRC) and internal model control (IMC) theory, an active disturbance rejection speed controller and current controller of the PMSM is proposed and designed. As the active disturbance rejection controller could observe the external disturbance, so the system is robust to the load torque variations. While the motor parameter variations observer is used for the compensation of the internal model controller, which could lessen the influence of the parameter variations. The whole vector controlled PMSM is robust to motor parameters and load torque variations.5. A sliding mode observer combined with phase-locked loop (PLL) and Extended Kalman Filter are designed for the speed estimation of PMSM, the estimated performance are also analyzed. Then, based on conventional MRAS, an improved MRAS stator resistance and speed estimated simultaneously is proposed, which could make the sensorless method have better speed estimation performance. And the proposed speed estimator and an integral sliding mode speed controller is combined for the sensorless control of PMSM at low speeds. At last, by combination of sliding mode variable structure and conventional MRAS, a new speed estimation method with quick response ability is designed. At the same time, an integal backstepping controller is designed for the speed control of PMSM, which has great robustness. The designed speed estimator with integal backstepping controller is combined for the sensorless of PMSM, which has good speed estimation performance and great robustness to external disturbances.