| Mineral processing is a process of beneficiating valuable minerals from underground mining of the ore. Production indices are very important to non-renewable mineral raw materials resource utilization efficiency, production quality and economic benefits of enterprises for mineral processing. It is difficult to ensure the production indicators are all in optimal conditions with manual operation, because the working conditions change frequently, the parameters are time-varying and the key production indices cannot be real-time online detected in ore dressing process. According to the production indices multi-objective optimization and considering the effect of varied equipment condition to the model, it provides a powerful tool for the production indices optimization and decision-making for mineral processing, which is of great research significance and practical value. This thesis takes the mineral processing of a large hematite concentration plant as background, and which is supported by national basic research and development program of China(973 Program,NO.2009CB320604) undertaken by the State Key Laboratory of Synthetical Automation for Process Industries. Research and development a production indices multi-objective optimization and decision-making algorithm for mineral processing, which combines with the object of production indices optimization for the ore dressing process to solve the problem of the production management and decision-making process. It is verified that the algorithm is correct and effectives through experiments. The main contents of the thesis are as follows:(1) First, this paper analyzed the problem of comprehensive production indices decision, and then give a description the problem of comprehensive production indices decision when device capabilities changing. Comprehensive production indices include concentrate grade, mental recovery, concentrate output, concentration ratio, and cost indicators. The method of production index optimization decision is to maximum mental recovery, concentrate output, concentrate grade and minimum concentration ratio, and cost indicators in the constrains of limited capacity constraints. As the actual production equipment capacity is not fixed, when the device capability changes, the production index optimization equipment capacity constraints will vary with changes in the actual situation.(2) Existing gradient-based drive hybrid evolutionary algorithm (G-NSGA-â…¡) is mainly to solve the static multi-objective optimization and is difficult to apply for dynamic problems. This paper put forward an improved dynamic multi-objective optimization algorithm based on initialization immigrants. First, this policy adding the random initialization immigrant populations, and enhance the diversity of population. Second, Calculate fitness function after the change occurs for the second parent and offspring, using the fitness value of the new evolutionary computation to get dynamic multi-objective optimization problem optimal solution, we use the standard test functions CONSTR and SRN with two constraints changing to simulate the beneficiation production indices optimization model, and has carried on the experiment test. Through the solution set contrast of the original G-NSGA-II algorithm and the new algorithm in the dynamic optimization problem, we verified the effectiveness of the algorithm.(3)We proposed production index optimization method in mineral processing equipment capacity changes, based on the above proposed immigration policy hybrid evolutionary algorithm and the features of production index optimization problems. We use actual data of a concentrator to experimental study for new constraint conditions change monthly integrated production index optimization decision problem. Experimental results show that the proposed algorithm can follow the changes of the environment, and get the new environment optimal solution. |
PDF Full Text Request