Research On Dynamic Optimization Of Industrial Process

From proper point, industrial process is a dynamic process, which means that the process state variables (such as concentration, temperature, pressure, flux and liquid-level etc) change with the change of time and space. Industrial dynamic processes are modeled with differential-algebraic equations (DAEs), where the DAE formulation consists of differential equations to describe the dynamic behavior of the system, such as mass and energy balances, and algebraic equations to ensure physical and thermodynamic relations.Over the past decade, applications in dynamic simulation have increased significantly in the process industries. These are driven by strong competitive markets faced by operating companies along with tighter specifications on process performance and regulatory limits. Moreover, the development of powerful commercial modeling tools for dynamic simulation, such as ASPEN Custom Modeler and gPROMS, has led to their introduction in industry alongside their widely used steady state counterparts. Unfortunately, up to know, little literatures give research on this area.One of the key issues of dynamic optimization problems is the solution algorithms. This paper is, therefore, to explore algorithm of dynamic optimization. The main work and contributions are as follows:1. The concept of dynamic optimization and it's application in chemical process are introduced. The development and the current application of algorithm of dynamic optimization are summarized. The advantage and disadvantage of these algorithms are further analyzed.2. The traditional control parameterization method is developed and two practical examples are illustrated. Based on the practical example analysis, the disadvantage of traditional control parameterization method is pointed out. An improved control parameterization method is then proposed, which combines two circulation method to improve the computational effect and iterative times with the approach precision guranteed simultaneously.3. Based on the analysis of control parameterization method, a novel algorithm is then propsed, where the variational method and two-point gradient are integrated.The results about the effect of the proposed algorithm on dynamic optimization problem are figured out in several classical examples.In order to improve the accurate of this algorithm , a adaptive -step algorithm is further proposed.4. For dynamic optimization problems with fixed boundary, a novel strategy named as objective-first is explored. Researches on several classical practical industrial examples are carried out, the results show the effect of the algorithm on practical industrial dynamic optimization problem.