Comparative Performance Analysis of PI Controller and MRAC for a Differential Drive Robot

This paper compares the performance of two control techniques on a DC motor-driven differential drive robot. Regular cascaded PI controller is the first scheme. In the second scheme, the outer PI controller is replaced by direct model reference adaptive control (MRAC). The main objective of the motor controller is that it follows a desired angular speed. Both schemes are compared in terms of average absolute speed error and power consumption. Preliminary simulations are performed on a DC motor model; whereas, experimental runs are done on a differentially driven robot powered by DC motors, to compare the performance of both control laws. It is found that the control scheme with MRAC has lesser speed error and power consumption compared to the PI controller, both in simulation and the experimental investigation.