Optimization of parameter adaptation algorithm in MRAS
Petar Crnošija, Željko Ban, N. Marijan · 2002
Application of a simplified parameter adaptation algorithm, which uses a total error of two state variables, has been considered. In this way, a total number of algorithm parameters was reduced to six. Parameter optimization used an integral criterion containing the square error weight with exponential function. By optimizing only two parameters, and keeping the values of other parameters constant, a satisfactory error and speed of adaptation was obtained for a large change of system parameters.>