Model Reference Adaptive Control using Multiple Controllers & Switching

In this paper, we develop an adaptive multi controller system to solve the MRAC problem using unfalsified control theory. Switching is done among the candidate controllers based on some suitably defined performance index, which is obtained without actually inserting the candidate controllers in the feedback loop. Though prior knowledge about the nominal plant structure or its parameters is useful to select the initial set of candidate controllers, this method makes no use nor tries to identify the plant structure or its parameters while deciding the optimal switching sequence among the candidate controllers. Simulation results are presented to indicate successful switching strategy. We also compare this approach with the switching scheme developed by Narendra for the MRAC problem in [2,7].