PERFORMANCE EVALUATION OF CLOSED LOOP DC MOTOR DRIVE

Abstract

DC motors provide speed control over a wide range with high starting torque plays significant role in modern industrial drives. Variable speed applications are being dominated by dc drives because of lower cost, reliability and simple control. Phase controlled rectifier drives faces the problem of having ripple content in output and control of PWM controlled rectifier drives is complex. Nowadays, the use of-low-inductance DC motors is becoming attractive because of the advantages of higher power density, lower rotor inertia, and smoother commutation and lower manufacturing costs but suffers from high ripple content, resulting in increased losses and switching losses. To overcome this, high switching frequency (order of kilo Hz) power converters are to be used. PWM converters cannot be used as they impose more switching losses and stresses. So, resonant converters are introduced which have high switching losses with a feature of reducing main switch switching losses. There are different types of error processing controllers such as PI, fuzzy, error limit control etc. where P1 is most commonly used controller because of it simple and robust but due to non-linearities in a DC motor it is difficult to apply conventional control algorithm for the speed control. Generally an accurate nonlinear model for the motor is difficult to find and the parameter values obtained from system identification may be only approximate values. So fuzzy logic, a perfect tool for modeling non linear equations which do not require a precise model is used to replace conventional control techniques is used for the speed control of DC motor. This Dissertation work involves performance evaluation between different types of converter fed DC drives using different controllers and comparison done among them. Simulation models are developed in MATLAB for different drive systems to compare and to evaluate the performance. Drive systems with two control techniques are introduced namely open closed loop system using NN model (reduces switching losses) and fuzzy controlled system (improves performance) along with error limit controller like PI and FUZZY controllers in my work.