On-line Optimization Of Continuous Industrial Processes

Real-time on-line optimization(RTOPT) plays an important role in the computer integrated automation system for industrial processes. With continuously monitoring the operation conditions and adjusting the steady state setpoints accordingly, the whole process is always driven to obtain the largest economic profits. RTOPT bridges the gap between basic process control systems and plant management systems that provide planning and scheduling capabilities.A lot of detailed techniques are needed for the implementation of on-line optimization, such as data reconciliation, parameter estimation, steady-state detection. On the other hand, some front ends and unsolved open problems in the optimization field are involved when optimization of the whole flowsheet is cncerned, such as optimization of large-scale process system, differential-algebraic optimization problems. Systematic research is carried out in this dissertation in an attempt to solve some of these important problems, and some new and interesting ideas are provided.●The overall structure of on-line optimization software for process systems is proposed. History of research works and industrial implementations of RTOPT is reviewed. Based on the analysis of the state of the art techniques involved in RTOPT, it is pointed out that industrialization and commercialization of RTOPT softwares is necessary and urgent.●The basic idea of open equation method for modeling and optimization is discussed. With open equation method, steady-state process models are constructed and all kinds of problems...