A metaheuristic approach for controller design of multivariable processes

A heuristic control design scheme is proposed for multivariable nonlinear processes. The control design problem is considered to be formulated as an optimization under constraints problem. Actuator, state and/or output variable constraints may be considered. The proposed heuristic control design scheme is based on a metaheuristic search algorithm, that utilizes process simulation blocks in a black-box approach. At each iteration of the algorithm new candidate controllers are randomly selected within the controller parameter space. The location and the range of the search area changes adaptively during the iterations of the algorithm.