Dynamic predictive planning+fuzzy logic control=intelligent control

The authors present a hierarchical, multiresolutional approach to control system design that addresses the problem of controlling the process in a logical, top-down, piecewise fashion. At higher (low-resolution) levels linguistic (symbolic) representation prevails, whereas at lower (high-resolution) levels analytical models remain the common tool of representation. For control based on a hierarchical description of the process, control algorithms or algorithm implementations need to be employed at the appropriate levels to account for the varying nature of the knowledge and the increasing performance requirements at higher resolution. These control schemes must map closely onto the hierarchical nature of the process description. The authors describe how dynamic predictive planning and fuzzy logic control can be combined to provide intelligent control of a dynamic, nonlinear process on the basis of this type of hierarchical framework