RBFNN-based multiple steady states controller for nonlinear system and its application

On-line Radial Basis Function (RBF) neural network based multiple steady states controller for nonlinear system is presented. The unsafe upper steady states can be prevented with the optimizer for Constrained General Model Controller (CGMC). Process simulator package is used to generate a wide range of operation data and the dynamic simulator is built as the real plant. The effectiveness is illustrated with a Continuous Stirred Tank Reactor (CSTR) and OPC tools are developed for on-line data acquisition and computation.