Research On Dynamic Optimization And Application In The Industrial Process

As the increase of requirement in the science, technology and industrialmanufactures, the traditional optimization methods often can not get the satisfyingperformance on the optimal control for non-linear, multi-dimensional and uncertaincomplex dynamic industrial processes. Therefore, dynamic optimization becomes agreatest important tool in chemical engineering design, optimal control and operationfor the industry. This paper pays main attention to the modification and improvement ofthe dynamic optimization algorithm.The detailed comparative researches among theproposed methods and other methods of Ph.D. dissertation of London university andgPROMs are carried out. This paper mainly includes the followings:1. Based on the control parameterization method, an improved algorithmemploying the idea of bilevel programming is proposed. Three classical casesare studied for examining the effective of the algorithm. And the researchresults show the effectivity of the proposed method.2. Orthogonal collocation method is always a main focus for the study ofdynamic optimization. A study is made to master principle and characteristicsof this method in a case of chemical engineering process. And a modificationis made in the collocation of control variable; then several simulations arecarried out to prove the effectivity of the modified method.3. Aimming at global optimization of the flexible and nonlinear chemicalengineering process, a thorough research is made on iterative dynamicprogramming (IDP) method. A new method to the adaptive dynamicprogramming is proposed to improve the main disadvantages of IDP, whichtakes a large computation time to excute optimization algorithm. The researchresults prove the effectivity of the proposed method by three classicalchemical engineering dynamic optimization cases.