Torque Control with Adaptive Fuzzy Logic Compensator for Permanent Magnet Synchronous Motor Chalermpon Pewmaikam 1+ , Jiraphon Srisertpol 2 , Chanyut Khajorntraidet 3 1 School of Mechanical Engineering Institute of Engineering Abstract. A torque control system of Permanent Magnet Synchronous Motor (PMSM) is an important process in industries, especially in the hard disk drive assembly processes. For example, the automatic screw machine is one of important machines in the hard disk drive assembly processes. The PMSM is a fundamental component of the automatic screw machine. The feedback torque control system use estimated torque received from motor current. The disturbance torque caused by the screw process depends on the quality of the screw heads and screw holes. The automatic screw machine requires precise output torque because the error of output torque affects the automatic screw machine performance. The damage in the hard disk drive assembly process may result from an inaccurate output torque. For instance, the defect of product is broken threads of the screws and the out of length of the screw heads. When the control process requires precision output torque, the machines must be calibrated by operator with standard value consumed 20-30 minutes per machine. This paper presents a torque control system with an adaptive fuzzy logic compensator for torque control and torque estimation simultaneously. The method of the research can increase the efficiency of torque control system and decrease the calibration time of the automatic screw machines. Keywords: Permanent Magnet Synchronous Motor, Control theory, Observer and Adaptive fuzzy logic compensator. 1. Introduction The screw driving process, one of the important machines in hard disk drive industry, can take part of hard disk drive to adjoin. By using a feedback control system of the automatic screw machine, torque of the machine converted from current and Linear Variable Differential Transformer (LVDT) is used to verify length of the screw heads. A screw driving process requires precise output torque. Many problems happen in the screw machine process, for example the screw and screw groove do not match and a quality of screw groove out of standard. The machines have to increase an output torque affected the physical system of the machine. Because of this reason, the physical system is changed. Each screw machine has different values of PID-controller parameters that improper for screw driving process. The screw driver of automatic screw must be calibrated by standard machine, which takes time in calibration process about 20 – 30 minutes per machine that affects the hard disk drive assembly processes performance. The dynamic responses of the PMSM motor transformed to the estimated rotor frame are nonlinear, thus the observer and observer error dynamics are nonlinear. The stability of control system is analyzed as a linearized error model [1]. A novel self-tuning PI controller is real-time designed according to the identified parameters based on pole assignment theory; the least-square estimator and a torque observer are used in the system [2]. A Fuzzy logic has been employed to control temperature. The error-count is used to trigger the fuzzy inference process [3]. A fuzzy PI-Controller is described that take into account some of the unique characteristic of such a furnace. Entries in the rule base are used to prevent integrator windup, and a fuzzy + Corresponding author. E-mail address:. chanyut_k@hotmail.com 31 2012 International Conference on System Modeling and Optimization (ICSMO 2012) IPCSIT vol. 23 (2012) © (2012) IACSIT Press, Singapore