Application of stochastic filtering to a permanent magnet synchronous motor drive system without electromechanical sensors

The Permanent Magnet Synchronous Motor (PMSM) is finding a wide range of applications as an ac servo drive in the motion control industry. Vector control is normally applied to the motor to convert it performancewise into an equivalent separately excited dc machine, which has highly desirable characteristics.; For a PMSM drive system, the instantaneous rotor position is needed to control the stator phase currents and the produced torque. The measured position is used also for position control or to derive the rotor velocity needed for speed control. This is normally achieved by using absolute position encoders or resolvers connected to the motor shaft.; In this project, the removal of this electromechanical sensor is considered. A stochastic filtering technique is presented to estimate the speed and the rotor position by only using measurements of the stator voltages and currents.; The estimator is an extended Kalman filter algorithm which is suitable to be implemented with a digital signal processor for on-line operation. It appears to be a viable and computationally efficient candidate for the on-line estimation of the speed and rotor position, since it accounts for all error sources acting on the system.; The mathematical process modeling of the motor and the structure of the filter algorithm are derived. The practical aspects of implementation are discussed. Simulation and experimental results of the overall drive system without a position sensor are carried out using the TMS 320C25 Digital Signal Processor to demonstrate the feasibility of this approach.; The main goals accomplished in this thesis are as follows: (1) Application of the Kalman filtering technique for the state estimation of a Permanent Magnet Synchronous Motor. (2) Design and development of the Kalman filter algorithm with a Digital Signal Processor. (3) Laboratory implementation of a four-quadrant torque/speed controlled, sensorless drive system with the DSP-based Kalman filter. (4) Verification of the proposed approach by simulation and experimental investigation. Suggestions for further research work are included.