Sensorless Adaptive Fuzzy Logic Control Of Permanent Magnet Synchronous Motor

• Main Author: Hafz Nour, Mutasim Ibrahim
• Format: Thesis
• Language: English; English
• Published: 2008
• Online Access: http://psasir.upm.edu.my/id/eprint/5344/1/FK_2008_4.pdf; http://psasir.upm.edu.my/id/eprint/5344/

Permanent Magnet Synchronous Motors (PMSM) require an electromechanical rotor position sensor to operate. The rotor position sensor has disadvantages, such as reliability, size, higher cost, and increased electrical connections. PMSM is used in many speed and position control industrial applications. Proportional integral (PI) and proportional integral derivative (PID) controllers have been widely utilised as speed controllers in PMSM drives. However, these controllers are very sensitive to step change of command speed, parameter variations and load disturbance. In this work, an adaptive fuzzy logic speed controller is proposed. The main features of the proposed controller are; quick recovery of motor’s speed from load disturbances and insensitivity to parameter variation over a wide speed range. The proposed controller is a hybrid model reference adaptive speed controller (HMRASC) which mainly consists of two functional blocks. The first block is a direct FLC that has the error and the change of error as inputs. The error signal is measured between the actual motor speed and the desired speed and the output is the change in the torque command. The second block implements a model reference adaptive controller. In the proposed system, the output speed of the reference model is compared with the actual speed of the motor and the resulted speed error is applied to a PI controller. The output signal of the PI controller is added to the direct FLC output to compensate any deviations in the motor speed from the reference speed due to parameters variation and disturbances in the load. The design and optimisation of the FLC are carried out using an adaptive fuzzy inference system network that uses the backpropagation, least square and gradient algorithms. The fuzzy inference system is trained and designed using an adaptive network. The rules and the implication method used are also optimised and minimised in order to shorten the computation time. In addition, the effect of different types and distributions of the membership functions were investigated and presented. This work also presents the estimation of the rotor position, which works effectively with nearly zero estimation error over wide speed range, to replace the electrometrical rotor position sensor. An estimation method based on the back EMF and flux estimation is presented to calculate the rotor position for medium to high speed. At low speed, the rotor position is calculated using signal injection where a high frequency low voltage signal is injected on the stator winding. In the proposed method, the measured motor’s current and the estimated motor’s voltage are processed through a signal processing block and a PI regulator to calculate the angle of the rotor position.Finally the performance of the HMRASC and the rotor position angle estimation algorithms are evaluated by simulation and verified experimentally for two motors using MCK2407 kit and IMDM15 board which are based on the TMS320LF2407 fixed point Digital Signal Processor (DSP) for different operating conditions. The first motor is rated at 50W and the second is rated at 380W. Both experimental and simulation results obtained from the HMRASC and the position angle estimation algorithms showed superior results compared to other methods presented in the literature.