Structure Adaptive Control Methods

The traditional adaptive control method is used for the system that the parametersare fixed or time-varied slowly, but the system is always contains some indeterminatefactors actually, so the structure and parameters of the system model changed with time.When there are many indeterminate factors and they are changed continually, theidentify part of the adaptive control method can't follow the change real time, whichmakes the system model inaccurate, so makes the control result unsatisfied. In view ofthese defects the Structure Adaptive Control method is proposed.This article applies the multi-model thought into the Structure Adaptive Controlmethod, and using the massive simulation experiments to confirm that the method isfeasible effective, the main results and innovations are indicated as follows:1. Proposed a method that using the hypothesis test approach to choose thesystem model. Elaborating its feasible foundation, then give the concrete hypothesis testapproach to the linear and the nonlinear system separately.2. Applied the markov process in the Structure Adaptive Control method. Usingthe markov transition probability and interactive multi-model algorithm to computeprobability that switch to each sub-model in current time, and carrying on the model cutaccording to it. Then, introducing the markov forecast model to forecast model changetendency real-time.3. Suppress the cut vibration using interpolation principle. In view of thevibration question which model cut brings, in the model forecast foundation, using theinterpolation principle to deal with the control quantity, so to guarantee the steadiness oftransient process.