Research On Intelligent Control Methods Of Electroslag Remelting Process

Electroslag remelting (ESR) process is a complex nonlinear system with inertia, multivariable coupling, time-delay and parameter change. These uncertainty factors make it difficult to perform the control task of total electroslag remelting process effectively by using conventional control methods. The intelligent control theory provides an efficient approach to realize the control of this kind of complex systems. Based on the mechanism and technique of electroslag remelting furnace, an intelligent control strategy is proposed in this dissertation by comprehensive utilization of mathematical model and intelligent control methods. The strategy includes the a multivariable decoupling control model of electroslag remelting systems, an intelligent optimal setting control strategy based on rough set and case-based reasoning (CBR) method, and a self-tuning PID compound control strategy of ESR current and voltage on different phases. The following research has been carried in this dissertation.The dissertation analyses the influence of the melting rate and the distance between poles on the quality of slag based on technique of ESR process. According to the dynamic characteristics of ESR, a calculation method is educed about the speed of drawing ingot, the rate of decline electrode, height of ingot and length of electrode. Experiments of set-point of ESR voltage and current are carried on and a mathematical model of constant slag resistance process is obtained in the dissertation.An intelligent optimal setting control strategy based on rough set and CBR method for controlling the ESR process was proposed according to the ESR technics characteristic, operation experiences and history database. The proposed control strategy optimizes ESR voltage and current set-points based on the integrated production process indicators under target, the boundary conditions and operating states of ESR process. The case search, revision and storage method in the CBR process are illustrated in detail.Aiming at the multivariable coupling characteristics of ESR process, an intelligent control strategy is proposed based on ESR mathematical model. It consists of multivariable decoupling controller, intelligent PID controller and the fuzzy self-tuning of the time delay. The intelligent PID controller parameters are optimized by Harmony Search based Particle Swarm Optimization (HS-PSO) algorithm and the fuzzy algorithm. A complex control structure of PID controller is proposed on different phase of ESR control. It actualizes self-tuning of PID on line, and perfects self-adjusting capability of PID control. The simulation research and industrial application experiments show that the model and the proposed control method are effective.The ESR production process utilized Siemens Profibus-DP fieldbus and industrial Ethernet technology to constitute the distributed control system in order to realize the centralized management and decentralized control on the ESR process. Finally, the intelligent modeling and control methods are applied to an ESR industrial production process and the application results are discussed. The results show that the control precision is improved from 10% to 15% of conventional control to 1% to 3% of the proposed control strategy. The system response time, tracking performance and stability are improved. The reseatch results are useful in improving the overall economic benefits of ESR enterprises and also provide a useful means for optimal control of complex industrial processes.