Study and analysis of DMC and application in D.C. motor speed control

Abstract

This project deals with the study, design and application of Dynamic Matrix Control for an unconstrained single input and single output system. The step response of the plant has been used for model prediction. The objective function used is “Least square objective function” which is based on squares of difference between predicted output and actual plant input .The strategy applied in DMC is to optimize the objective function, for this we use a unique expression of feedback matrix k (which is obtained through simple calculus). Further the response of plant being controlled by DMC controller is obtained which are significantly better than the step response of the plant. Then study of the effect of DMC parameters P (prediction horizon), N (model length), M (control horizon length), W1(error weight matrix), W2(control weight matrix) on plant response have been done. After this, effect of disturbance on the plant response and effect of weighting tuning parameters is also discussed. A plant with random noise and it’s control is also discussed. The last chapter of this thesis holds the application domain, which shows the applicability of DMC through a simple DC motor. First, modelling of DC motor is done, then it is further tested in SIMULINK. Then the same is used for getting appropriate set-point trajectory of motor output at a set armature voltage for no load case. Further in this section, control of motor with no load and with load, both the cases are discussed. Further, capability of DMC with different load values is tested. Next comes testing DMC control law against varying load and then DMC is tested against variation of loading time. In all the above given scenarios DMC performs very well. All the above analysis is supported by MATLAB simulations.