Speed Control of BLDC Motor By Using PID Control and Self-tuning Fuzzy PID Controller Mohammed Abdelbar Shamseldin Mechatronics Engineering Department Future University at Egypt Cairo, Egypt eng.moh.abd88@gmail.com Adel A. EL-Samahy Electrical Power and Machines Department Helwan University Cairo, Egypt el_samahya@yahoo.com Abstract— This paper presents three different robust controller techniques for high performance brushless DC (BLDC) motor. The purpose is to test the ability of each control technique to force the rotor to follow a preselected speed/position track. This objective should be achieved regardless the parameter variations, and external disturbances. The first technique is conventional PID controller. The second controller technique use genetic algorithm to adjust the PID controller parameters based on three different cost functions. Finally a self-tuning fuzzy PID controller is developed and tested. These controllers are tested for both speed regulation and speed tracking. Results shows that the proposed self-tuning fuzzy PID controller has better performance. Keywords—PID control; Self-tuning fuzzy PID control; Genetic algorithm; I. INTRODUCTION The development of high performance motor drives is very important in industrial as well as other purpose applications such as steel rolling mills, electric trains, electric automotive, aviation and robotics [1]. Several types of electric motors have been proposed for these types of applications [2]. Among these types, conventional DC motors which are known with their excellent characteristics. On other hand conventional DC motor, have some disadvantages such as routine maintenance of commutators, frequent periodic replacement of brushes and high initial cost [3]. Conventional DC motors cannot be used in clean or explosive environment. Squirrel cage induction motor is alternative to the conventional DC motor. It offers the robustness with low cost. However, its disadvantages have poor starting torque and low power factor [4]. In addition, neither conventional DC motors nor induction motors can be used for high-speed application. The alternative to both conventional DC motor and induction motor is the DC brushless motor, which can be considered the most dominant electric motor for these applications [4]. They are driven by dc voltage but current commutation is done by solid-state switches. The commutation instants are determined by the rotor position and the position of rotor is detected by position sensors or by sensorless techniques [4]. The state space model of BLDC motor is discussed in [4]. This model derived for BLDC motor in arbitrarily reference frame. One of the major drawbacks of such a model is the dependency of the state variables coefficients on the rotor position (time dependent), which make difficult to implement [5]. This problem may be alleviated by considering the model presented in [5] and adopted in this paper. The coefficients of state variables using this model are constant with time. This is the main advantage of such a model. In high performance drives applications such as robotics, dynamic actuation and guided manipulation, moving the end effector from one position to other is not the only objective, the end effector while traveling must follow a pre-selected time tagged trajectory at all times [1]. This must be achieved even when the system loads, inertia, and parameters are varying, to achieve this objective the control strategy must be adaptive, robust, accurate and simple to implement [1-4]. The PID controller is applied in various fields in engineering owing to its simplicity, robustness, reliability and easy tuning parameters [1]. The famous method to find PID parameters Ziegler-Nichols rule but sometimes are not the best. So, it can be realized by using genetic optimization technique based on three different cost functions to find the best PID control parameters. The main obstacles facing PID control technique is sudden change in set point and parameter variation, it makes the PID control gives poor response [4]. This problem can be alleviated by implementing advanced control techniques such as adaptive control, variable structure control, fuzzy control and neural network. One of the major problems with implementation of self-tuning adaptive control techniques is inability to achieve the trajectory control in the presence of sudden disturbances or large noises. This is because the parameter estimator might provide erroneous results in the presence of sudden disturbances or large noises [6]. Variable structure controller is simple but it difficult to implement. This is because of the possibility of the abrupt change in the control signal, which might affect the system operation [2]. A neural-network-based motor control system has a strong ability to solve the structure uncertainty and the disturbance of the system, whereas it requires more computing capacity and data storage space [4]. Fuzzy control theory usually provides non-linear controllers that are capable of performing different complex non-linear control action even for uncertain non-linear systems [1]. Unlike conventional control designing, a FLC does not require precise knowledge of the system model such as the poles and zeroes of the system transfer function [1]. A fuzzy-logic control system based on expert knowledge database needs less calculations, but it lacks sufficient capacity for the new rules [4]. So, the combination between fuzzy and PID controller for tuning the PID parameters according to the error and change of error might be a good alternative. It is 15th International Workshop on Research and Education in Mechatronics (REM), Elgouna, Egypt, September 9-11, 2014 978-1-4799-3029-6/14/$31.00 ©2014 IEEE