The Minimum Time Dynamic Optimization Of TE Process Grade Transition

Static optimization techniques assume industrial operation process is a steady state; however, the actual production processes which have significantly dynamic characteristics are difficult to satisfy this assumption. In particular, the development of modern chemical trend is meticulous production. It is characterized by:the process of intermittent, time-varying, nonlinear model, the product requires a high-purity, small quantity and variety. Steady-state process in a complex chemical process is only a relatively temporary, in fact, the actual process are full with fluctuations, interference and condition variables. Because these processes essentially are dynamic, static methods have been unable to solve these problems. It is very important to develop the corresponding dynamic simulation and dynamic optimization method.For dynamic optimization problem, the terminal times given the issues being studied are more common, and for the terminal time not given the problems studied rare. The focus of this research is not given terminal time dynamic optimization problem.According to the characteristic of minimum time dynamic optimization problems, a double-layer optimization algorithm (DLOA) is proposed to solve these problems. The inner optimization is to construct optimal control problem with free final states. Differential evolution algorithm is used to find the optimal solution in given terminal time. In the outer, DLOA calculated the time range of next iteration according to the inner calculation. This method is particularly suitable for the mechanism of the model is not clear or the study of the black-box structure. When applied to typical minimum time dynamic optimization problem, DLOA demonstrated a competitive optimal searching ability and more accurate optimization results.According to the characteristics of Tennessee-Eastman (TE) process grade transition, combined with the general strategy proposed in this paper to solve the minimum time problem, introduced the process dynamically adjust prior knowledge, this article further propose targeted solutions which is called’proposed evolutionary programming method based on the incremental control variable trajectory characteristics’. This method improves the randomness in the process optimization, enhances optimization efficiency and implementation. By comparing the TE process result using artificial experience in the operation and the result of optimization trajectory, it demonstrated that the dynamic optimization method can effectively guide the operation of the production process operations, significantly improve system security and reliability.