Intelligent control system design

Process control is implemented at various levels in the manufacturing process industries. At the micro level it deals with implementing safety interlocks, tuning controllers, positioning actuators. etc. At the textit macro level it deals with plant wide optimization and control, supervisory control and scheduling. The micro level implementation of process control uses programmable logic controllers (PLC) and dis- tributed control systems (DCS) The macro level implementation uses workstations and main frame computers with communication networks that talk to the controllers at the micro level. Closed analytic representation is unavailable at the macro level, and sometimes at the micro level. Artificial Intelligence plays an important role in implementing process control at the micro and macro level. This research looks at implementing Intelligent Control on an instrumented shell and tube heat exchanger in the Unit Operations laboratory. The heat exchanger is controlled by an industrial controller (PLC) and an operator interface. System Cultivation tools are developed to monitor and classify process variable trends. These shallow reasoning tools help determine if a process is at steady state, under transition or oscillating. Information from these tools is used by a expert system to provide intelligent supervisory control and intelligent system identification. A framework has been developed to provide an easy way to implement new control strategies and system identification techniques which when combined with the intelligent features of the controller provide a versatile and fail safe controller.