Font Size: a A A

Copper Flash Smelting Ingredients Research And Software Development Plan Optimization Algorithm

Posted on:2013-12-20Degree:MasterType:Thesis
Country:ChinaCandidate:J C LiFull Text:PDF
GTID:2241330377953471Subject:Non-ferrous metallurgy
Abstract/Summary:PDF Full Text Request
The ingredient of copper flash smelting is a key step to copper smelting and production. And the primary task of ingredient is to make the ingredient plan. A scientific and proper ingredient plan is of great significance to the stable flash furnace production and high-quality products. Meanwhile, it should exert positive on the establishment and planning of enterprises’ production planning. Most of the raw materials of copper smelting are copper ore concentrates imported from foreign countries, which are of various kinds, unstable, and are not all of good quality, resulting in a frequent changing of ingredient designs. As the timeliness and optimization of the current man-made ingredient plan cannot meet the needs of the rapid development of enterprises, it has great study significance and practical value in research the mathematical models on the ingredient of copper flash smelting and developing the software for the optimization of computer-ingredient-plan.This paper focuses on the following researches by taking the ingredient of copper flash smelting and production in Tongling Nonferrous Metals Group Holding Co., Ltd. as the background:1. By doing an in-dept analysis on the influencing factors to ingredient optimization of copper flash smelting, and with the reference of the practical experience in the factory, an object function of maximum-use oriented ingredient optimization of copper flash smelting is put forward. Moreover, by taking into account the constraint condition on ingredient optimization, a mathematical model for the ingredient plan of copper flash smelting is established.2. Based on the models established, a research on the general algorithm is carried out. After massive calculations, a optimization algorithm for the ingredient and production of copper flash smelting, namely "Second Heredity--Enumeration Arrangement", is eventually brought forward.3. Make an improved version of genetic algorithm by proposing "No encoding-bandwidth variation". Giving full consideration to thee characteristics of this case, this algorithm retains the historically best individual by selective operation, cancels the code, the decoding and the interlace operation in the tradition genetic algorithm, controls the variation band width of various variables, in order to steer it toward the sound track. The large number of example calculations show that the algorithm has advantages of high target rate and high optimization degree in this case.4. This research selects the basic genetic algorithm to carry on the contrast inspection. Under the conditions of possessing same stock of minerals type, mineral content, size of population and evolution algebra, The degree of optimization of the basic genetic algorithm, algorithm of no coding-bandwidth variation were89.28%,95.58%respectively.5. Propose the probabilistic method to objectively evaluate the pros and cons of randomized algorithms. A genetic algorithm (including the particle swarm, ant colony algorithm) is a kind of randomized algorithm in essence. Due to randomness, any kind of stochastic algorithm results can not be effectively proved that the pros and cons of their algorithm only by a few calculated result. This study uses the massive counting result to calculate the mathematical expectations and standards deviations of each generation after evolution. So as to assess the pros and cons of randomized algorithms.This study explores the long-term ingredient plan of copper flash smelting through the computer instead of man-power, and achieved the expected effects; the results of ingredients can basically meet the requirements of the actual process of production, and get the aims to work out the ingredient plan by computer, providing a basic for the realization of the plan optimization of production and ingredients.
Keywords/Search Tags:Flash smelting, ingredient plan, genetic algorithm, non-coding heredity, bandwidth variation
PDF Full Text Request
Related items