The development of tuning techniques for single-input single-output and multivariable control systems using controller sensitivity measures

The development of efficient techniques for the on-site optimisation of controller parameters is a topic of considerable importance in control engineering. Controller tuning techniques which do not depend upon the availability of an accurate plant model or upon trial-and-error methods are of special importance. This paper describes simulation-based investigations of one possible approach to control system tuning which is applicable to continuous and digital controllers and can be extended readily to multivariable control systems.