Adaptive Regulators and Iterative Tuning

This chapter discusses the adaptive regulators and iterative tuning methods. The algorithms of recursive adaptive regulators are treated using the canonic forms of Tsypkin. The algorithms of adaptive dual control are presented; then a special class of these algorithms, the self-tuning regulators are shown. Simple, but very convincing simulation examples show the operation of these adaptive regulators. Another class of adaptive regulators is based on simultaneous identification and control, that is, repeated off-line closed-loop identification methods. The general forms of such methods in iterative schemes are presented together with convincing simulation experiments. It is also shown how the exciting reference signal can be optimized, providing the optimization of the modeling loss. The reference model redesign technique is then presented, which helps the optimization of the design loss. All these methodologies are supported by simulation examples. The triple control is the most complex version of adaptive control, where the adaptive dual control algorithm of the recursive process identification and regulator tuning is combined with the optimization of the exciting reference signal. Simulation examples prove how fast the learning adaptation process is in these cases.