Dissertation > Industrial Technology > Metallurgical Industry > Metallurgical Technology > Remelting > ESR Process

Design of Control System and Research of Control Method for Electroslag Furnace

Author GaoXinPing
Tutor QinShuKai
School Northeastern University
Course Control Theory and Control Engineering
Keywords electroslag furnace neural network model control method harmonysearch
Type Master's thesis
Year 2011
Downloads 8
Quotes 0
Download Dissertation

The technology of ESR is a new kind of smelting, which is developed in the1930s. Due to its huge advantage in purifying steel quality and improving alloy internal crystalline organization, it develops very fast. Because of the fast development of society, new and high quality alloy steel demand is increasing. Therefore, the study of effective ESR control system to increase productivity and the quality of products and energy conversation is of vital significance.Based on the transformation of control system of electroslag furnace, this paper uses Data-driven approach to establish a current prediction model, and introduces the prediction model into the controller, which will improve the performance of the controller. The main contents are as follows:1) After analyzing the structure and process of electroslag furnace and learning advanced electroslag furnace intelligent control system design experience home and abroad, this paper designs a practical control system with local conditions.2) Modeling for industrial furnace is an enduring topic. On the basis of field data collected back and RBF neural network technology, this paper establishes a current prediction model. To get a more accurate model, rival penalized competitive learning and improved PSO are introduced into RBF neural network’s weights optimization.3) Because of the traditional PID control unadaptable and parameters not easily changed, this paper proposes a new kind of control method, harmony search prediction of PID, which uses harmony search to optimize the three parameters in PID controller and gets the RBF prediction model’s output as its feedback value. Through the simulation contrast, it is concluded that the harmony search prediction PID is an excellent control method in theory.

Related Dissertations
More Dissertations