基于数学规划及神经网络的釉料配方优化研究

发布时间：2018-01-03 23:31

本文关键词：基于数学规划及神经网络的釉料配方优化研究　出处：《景德镇陶瓷大学》2016年硕士论文　论文类型：学位论文

【摘要】：釉料在陶瓷釉的生产中起着非常重要的意义,因为原料的成本在市场中会不断的变化,所以釉的生产商往往需要跟随市场的波动来调节釉的配方,用来确定釉料的成分配比,最后达到市场的最小值,以节约成本。由于釉在国内的生产量丰富,即使配方成本降得很低,也可以为釉制造商带来显著的经济利益,快速计算釉的最低成本配方,在釉行业的配方形成过程中,计算机在釉配方设计中有不可替代的地位和作用。用一切办法和手段来达到成分的最低要求,使得釉料的价格降到最低,满足最低成分要求,得到最优化配方,这是我们最终的研究目标。本文将运用四种数学规划方法,即:线性规划、目标规划、随机规划、模糊规划。同时还利用两种数学优化算法,即:模拟退火算法和遗传算法,研究釉料配方的优化设计,但是上面各种方法的都有自己的优点和缺点,选择什么样的方法由具体的市场情况、釉标准和釉料成分来共同决定,以及许多其他的因素。本文作者选择釉料的成分、釉料的配比、釉料的成本、釉料的国标,从以上四个方面考虑釉料的配方关系,最后利用人工神经网络(自组织特征映射(SOFM)研究各种釉料配方的优化决策,从而决定最好的釉料配方,由于自组织特征映射(SOFM)神经网络对这种配方优化研究有很强的适用性,因此根据其聚类结果可以选择最佳的配方。
[Abstract]:The glaze plays a very important role in the production of ceramic glaze, because the cost of raw materials will continue to change in the market, so the glaze manufacturers often need to follow the fluctuations of the market to adjust the glaze formula, used to determine the composition of the glaze, finally reaches the minimum value of the market, in order to save costs. Because in the glaze the amount of domestic production of the rich, even if the formula cost drop very low, also can bring significant economic benefits for the glaze manufacturers, the lowest cost formula of fast calculation of glaze, glaze industry in the formulation process, the computer has an irreplaceable role in the glaze formulation design. The minimum requirements for all the ways and means to achieve the composition, the glaze of the price down to a minimum, meet the minimum requirements of components, optimized formula, this is our final research goal. This paper will use four kinds of mathematical programming method, namely: linear programming Planning, programming, stochastic programming and fuzzy programming. At the same time using the two algorithms, namely, simulated annealing algorithm and genetic algorithm, the optimization design of the glaze formula, but the above various methods have their own advantages and disadvantages, the choice of what kind of method by the specific situation of the market standard glaze and glaze composition is determined, and many other factors. The author selected the glaze composition, glaze glaze proportion, the cost of the glaze glaze formula considering the relationship between the national standard, from the above four aspects, finally using artificial neural network (self-organizing feature mapping (SOFM) optimization of various glaze formula, glaze the best formula to decide, because of the self-organizing feature map (SOFM) neural network has very strong applicability of the formula optimization research, according to the clustering results to choose the best formula.

【学位授予单位】：景德镇陶瓷大学
【学位级别】：硕士
【学位授予年份】：2016
【分类号】：TQ174.43

【相似文献】