Dynamic Optimization And Its Application Based On IDP

With the development of industrial modernization, it is important to control the production process precisely, as it can both ensure the safety of the production process and improve economic efficiency, which promote the development of dynamic optimization techniques. In recent decades, dynamic optimization methods are emerging in an endless stream with good characteristics in different areas. But the system is more and more complicated with the reformation in industry. People raised higher requirements to the production process and performance indicators. Experts and scholars need to do research in improving the performance of dynamic optimization methods. The direct methods can optimize systems directly without gradient information, but they are slow and have low accuracy. The indirect methods can optimize fast with gradient information, but they are easy to get caught in local optimum. I consider combining the advantages of both dynamic optimization methods to optimize systems in this paper. Combined with the actual system development background, this paper discusses the Nuclear Power Plant steam generator level control process with the dynamic optimization method. This process’s special thermal-hydraulic environment is rather complicated and causes more difficulties to optimize the system. Finally this paper achieves a Nuclear Power Plant research and development of the monitoring system for the nuclear island. The main work of this paper is as follows:(1)This paper does in-depth research in iterative dynamic programming and first-order algorithm. With the process of achieving the two algorithms, I understand the advantages and disadvantages of them. This paper proposes an improved iterative dynamic programming algorithm. The algorithm learns from both of the iterative dynamic programming and first-order algorithm. The algorithm has the high accuracy and speed features which are learned from first-order algorithm. It also has the good global optimization feature which is learned from iterative dynamic programming algorithm. At last I use the proposed algorithm to two process optimization problems which are continuous stirred tank reactor(CSTR) process and fed-batch temperature control process. The results illustrate the effectiveness and advantages of the proposed algorithm.(2)This paper mainly studies for steam generator of Pressurized Water Reactor Nuclear Power Plant. I take an in-depth study in the work process and liquid level characteristic of steam generator. The main factors, which affect steam generator level, are stated. The complexity of steam generator level is presented. A classic level process simplified model proposed by predecessor is thought as the object of study. Through the analysis of this model and the requirements of the level control process, this paper gets the proposition of the dynamic process. The improved iterative dynamic programming algorithm is used to solve this proposition. I get a satisfied optimization result. Also it illustrates the applicability of the proposed optimization algorithm to this dynamic optimization problem.(3)The implementation and development of Nuclear Power Plant monitoring system which is based on real-time database is the background of this paper. I achieve part of the monitoring system. Firstly, I give the introduction of the real-time database and two methods of communication. Then, I build the database and accomplish the monitor interface. Through the description of the hardware firewall, I tell about the configuration of the network security. Lastly, I introduce the development of the steam generator control optimization module.