基于模拟退火遗传算法的变压器硅钢片优化排样的研究

发布时间：2018-06-29 02:12

本文选题：遗传算法 + 模拟退火　；参考：《吉林大学》2012年硕士论文

【摘要】：二维排样问题是一类具有最高计算复杂度的NPC（Non-deterministicPolynomial Complete）类问题，同计算机图形学、人工智能、神经网络、运筹学等学科有着广泛的联系与交叉，被广泛运用于服装、机械制造、玻璃、造纸等行业。研究排样问题不仅能给社会带来不可估量的经济价值和社会价值，而且能进一步促进社会向环境保护型、资源节约型方向发展。计算机技术的不断更新和发展，排样方式已由以前的手工排样发展到计算机的辅助排样，排样算法已由简单的搜索算法发展到智能的启发式算法，排样复杂度也随排样件的形状和计算复杂度的增加而增加。本文研究了变压器铁芯片的排样问题，主要工作和成果如下： 1.深入研究国内外排样问题的发展现状和趋势，以及变压器铁芯片现行的排样方式和局限性。在现在的变压器铁芯片制造基础上，根据变压器铁芯片的种类、形状和片数的特点提出了本文的排样系统，建立了排样系统框架，并且阐述了排样系统的基本内容和理论。 2.在本文的排样系统中，横向排样和纵向排样的系统是分开的。横向排料的目的是找出排入板材的零件种类，纵向排样的目的是找到参加排样的每个零件排在这张板中的片数，使得整张排料图的利用率达到要求。 3.横向排样的直接结果就是找到这样的零件排列组合，使得排入板材的零件的宽度和尽量接近板宽，，也就是这些零件排在板材上时，尽可能占满板材。横向方向排样时，采用遗传算法中融入模拟退火的算法的优化算法，克服了遗传算法早熟和优化结果收敛性不够好的缺点，使算法每次都能解出横向板材利用率最高的最优解，保证了每次整体排样中横向排样结果的最优，最大限度的节约了硅钢片。 4.在纵向方向排样时，根据动态规划理论中的“最优化原理”建立了排样算法的数学模型，一张排料图完成后，剩余的零件也用这种方法进行排样。这样就可以保证每次排样的结果都是最好的，到最后整体排样所有的排样图加在一起的板材利用率就一定是最好的。 5.本文在MATLAB上对这种排料方法进行了仿真，并研究了板材横向宽度值的选择与余料的关系以及纵向剪切长度范围选择与排料利用率的关系，得到了最合适的板材宽度和纵向剪切范围。
[Abstract]:Two-dimensional layout problem is a kind of NPC (Non-deterministic Polynomial complete) problem with the highest computational complexity. It has extensive connections and intersections with computer graphics, artificial intelligence, neural network, operational research and so on, and has been widely used in clothing, mechanical manufacturing, glass, etc. Papermaking and other industries. The study of layout can not only bring inestimable economic value and social value to society, but also further promote the development of society towards environmental protection and resource conservation. With the continuous updating and development of computer technology, the layout method has developed from the previous manual layout to the computer-aided layout, and the layout algorithm has developed from simple search algorithm to intelligent heuristic algorithm. The layout complexity also increases with the increase of the shape and computational complexity of the sample layout. In this paper, the layout of transformer core sheet is studied. The main work and results are as follows: 1. The present situation and trend of layout problem at home and abroad, and the current layout mode and limitation of transformer core sheet are studied in depth. Based on the manufacture of transformer core sheet, according to the characteristics of transformer core sheet type, shape and number of chips, the layout system of this paper is put forward, and the frame of layout system is established. And elaborated the layout system basic content and the theory. 2. In the layout system of this paper, the horizontal layout system and the longitudinal layout system are separated. The purpose of horizontal layout is to find out the type of parts that are placed into the plate, and the purpose of longitudinal layout is to find out the number of pieces of each part arranged in this board, which makes the utilization ratio of the whole layout diagram reach the requirement. 3. The direct result of horizontal layout is to find such parts arrangement and combination, so that the width of the parts into the plate and as close to the width of the board as possible, that is, when these parts are arranged on the plate, they occupy the full plate as much as possible. In transverse layout, the genetic algorithm with simulated annealing algorithm is adopted, which overcomes the shortcomings of premature convergence of genetic algorithm and poor convergence of optimization results, and makes the algorithm solve the optimal solution with the highest utilization ratio of transverse plate every time. It ensures the optimum result of transverse layout in each overall layout, and saves silicon steel sheet. 4. 4. In longitudinal layout, the mathematical model of layout algorithm is established according to the "optimization principle" in dynamic programming theory. After a layout diagram is completed, the remaining parts are also arranged by this method. This can ensure that the results of each layout are the best, to the end of the overall layout of all the layout chart added together the plate utilization ratio must be the best. 5. This paper simulates this method in MATLAB, and studies the relationship between the selection of transverse width of sheet and the residual material, and the relationship between the selection of longitudinal shear length range and the discharge utilization ratio. The most suitable width and longitudinal shear range are obtained.
【学位授予单位】：吉林大学
【学位级别】：硕士
【学位授予年份】：2012
【分类号】：TH186;TP391.72

【参考文献】