RESEARCH ON A SELF-TUNER WITH CHANGEABLE PERFORMANCE DURING OPERATION

A generalized minimal variance stochastic self-tuning controller with a changeable control performance criterion is proposed in this paper. The law which determines the weighting transfer functons in every sample interval is solved through the pole-Placement. This control algorithm is quite simple. Simulation results show that the strategy of variable weighting transfer functions is superior to the ordinary selftuning controllers. Finally, the results of the full-scale test on a steamer steering are given.