Research On Multi-objective Optimization Decision Method And Its Applications To Copper Electrolysis Process

Copper is a kind of very important raw materials and the basis of strategic goods. Electrolytic copper production process is the main approach to obtaining high purity refined copper products. However, Electrolysis production process has the characteristics of complex industrial process, such as the complex reaction mechanism, nonlinear, strong coupling and time-varying. The important parameters related to product quality are difficult to be measured on line. And the model between production goals and operating parameters is difficult to establish. These reasons affect the optimal control of electrolytic copper production process. Nowadays, along with the changes in market demand and the improving of product quality requirements, copper electrolysis need to be accelerated development using high-tech industrial. Based on the electrolytic copper production technology maturing and the level of process control system and equipment constantly improving, the parameters soft sensor and stable state optimization technology have become two important and closely linked studies to further improve the economic efficiency of enterprises.Based on the actual process of industrial production electrolytic copper, some problems and solution strategies in stable state optimization and decision method of the industry process are discussed. The researches include two major parts:The 1st part based on the analysis of factors affecting the quality optimization control problem and power consumption, considering the multi-objective constrained optimization problem in the common industrial process, a new multi-objective optimization is proposed to solve the steady-state optimization problem of copper electrolysis process. The 2nd part is the research on decision-making method. Based on the multiple attribute utility function, an improved method is proposed for the process control system stable state optimization. The main works can be summarized as follows:1. Based on the production of electrolytic copper principle, industrial production processes in-depth analysis, the soft sensor model of copper and acid ion concentration and auxiliary variables can be confirmed; the power consumption for the economic indicators and the copper and acid ion concentration for the quality indicators as the three optimization functions are set up.2. A new MOPSO is presented to solve this multi-objective optimization. Using variable size external set and fast sorting method to reduce computation time, the crowding-distance operator and enhanced dominance are also adopted to maintain the diversity and wide spread of solutions. The simulation the results and the applications in the electrolytic copper optimization of industrial process show that this proposed MOPSO has good quality in the computation time, diversity and wide spread of solutions, especially in the question of three objects optimization. So this method is more practicable in the multi-objective optimization.3. An improved multiple attribute utility function decision-making is proposed for the process control system stable state optimization. Through changing the decision variables (the set value of controller), the controlled process can be transferred from one stable state to another stable state. However, the dynamic responses go with this course. If using classic decision method, the set value of controller may be changed widely and inappropriately; that may influence the stability of system. In order to avoid this defect, the decision preferential information and the change of decision variable are all considered into this improved method to fit the stable state optimization system. The simulation the results and the applications in the electrolytic copper show that the new approach takes account of weight and decision variables. That can avoid influencing the stability of system, and maintain an appropriate degree of optimization process.Finally, we summarize all our works, put forward views of process optimization theory research and engineering applications, and the problems to be further studied are analyzed and viewed.