Dynamic systems identification and control is a very indispensable stage in application of practical control. This paper focuses on identification of a preferable technique for controlling the armature of a laboratory DC motor, rotating in one direction. The most important procedures for tracking set point commands and reducing sensitivity to disturbances which occur as a result of the load was investigated based on three different control methods. The motor Input-output data was attained offline using a MATLAB implemented recursive least square algorithm, a linear model transfer function having two poles-one zero is obtained, subsequently simulations in normal MATLAB program and Simulink are carried out to assess the model transient and steady state performances. The three techniques that were investigated are; feedforward, feed-backward (root locus) and Linear Quadruple Regulator (LQR) control techniques for DC motor control. Result obtained from the simulation showed that the influence of second control loop was obvious as it successfully locked the integral feedback control set-point and the target speed of the machine. The LQR method obtained a more precise stability and steady speed on close loop with respect to load. This was an indication the LQR compensator had a better performance during load disturbances rejection when compared with the other two techniques.
Key words: DC motor modelling, Load Torque Condition, LQR compensator, System identification, Tracking set point
|
