锻压技术

基于BLNFP神经网络算法的二维不规则钣金零件排样

英文标题：Layout for two-dimensional irregular sheet metal parts based on BL-NFP neural network algorithm

作者：刘玲玲赵罘龚堰珏

分类号：TH164

出版年,卷(期):页码：2021,46(12):54-60

摘要：

针对传统工业生产中二维不规则钣金零件的利用率不高、计算时间较长的问题，提出了以能量函数为载体的BLNFP（Bottom Left No Fit Polygon）神经网络算法。该算法是将BL（BottomLeft）定位算法和临界多边形（NoFit Polygon，NFP）几何特性相结合，同时模拟了钣金零件的排样过程。并采取对待排入零件优先进行面积大小核算、再排入待排物体的方式，并利用Matlab对算法的输出数据和前人所列出的数据结果进行了测试对比。结果表明：BL定位算法能够合理地计算出零件的排入位置，NFP能够有效地解决不规则零件排样利用率小的问题，神经网络算法则能够有效地提高求解速度。针对二维不规则钣金零件的排样问题，与传统神经网络算法相比较，采用该算法缩短了钣金零件下料机器计算最优解40%的时间，并提高了约8%的钣金材料利用率。

For the problems of low utilization rate and long calculation time of two-dimensional irregular sheet metal parts in traditional industrial production, a BL-NFP (Bottom Left-No Fit Polygon) neural network algorithm with energy function as the carrier was proposed, which combined BL (Bottom-Left) positioning algorithm with the geometric characteristics of critical polygon (No-Fit Polygon, NFP), and the layout process of sheet metal parts was simulated. Then, the method of calculating the area sizes of the parts to be placed first,and then placing the objects to be arranged was adopted. Furthermore, the output data of the algorithm and the data results listed by the predecessors were tested and compared by Matlab. The results show that BL positioning algorithm can reasonably calculate the placing position of parts, NFP can effectively solve the problem of low layout utilization rate for irregular parts, and the neural network algorithm can effectively improve the solution speed. Aiming at the layout problem of two-dimensional irregular sheet metal parts, compared with the traditional neural network algorithm, the time of the optimal solution calculated by blanking machine of sheet metal parts is shortened by 40% using this algorithm, and the utilization rate of sheet metal material is improved about 8%.

基金项目：

国家自然科学基金资助项目（51975006）

作者简介：刘玲玲（1995-），女，硕士研究生 E-mail：137757301@qq.com 通信作者：赵罘（1972-），男，博士，副教授 E-mail：zhaof@btbu.edu.cn

参考文献：

[1]高荣宇. 二维不规则件排样优化系统研究与实现[D]. 长沙：湖南大学, 2019.

Gao R Y. Research and Implementation of Two Dimensional Irregular Parts Layout Optimization System[D]. Changsha: Hunan University，2019.

[2]Wu D, Yan C.A balance approach for the onedimensional multiple stock size cutting stock problem with setup cost[J]. Proceedings of the Institution of Mechanical Engineers Part B Journal of Engineering Manufacture, 2016, 230(12):2182-2189.

[3]Plácido R. Pinheiro, Bonfim Amaro Júnior, Rommel D. Saraiva . A randomkey genetic algorithm for solving the nesting problem[J]. International Journal of Computer Integrated Manufacturing, 2016, 29(11):1159-1165.

[4]王静静, 瞿少成,李科林.一种基于并行交叉遗传算法的二维不规则排样问题求解[J].计算机应用与软件,2020,37(7):188-193.

Wang J J, Zhai S C, Li K L. A parallel crossover genetic algorithm for 2D irregular packing problem[J]. Computer Application and Software, 2020,37(7):188-193.

[5]Vasundara M, Padmanaban K. Optimization of fixture layout and artificial neural network (ANN) weights of ANNfinite element analysis based fixture layout model using genetic algorithm[J]. Journal of Engineering & Technology, 2014, 4(2):102-109.

[6]张娜, 赵罘.基于非等值初始量蚁群算法的矩形优化排样[J].北京化工大学学报:自然科学版, 2019, 46(6):72-77.

Zhang N, Zhao F. Optimization of a rectangular layout based on a nonequivalent initial pheromone ant colony algorithm[J]. Journal of Beijing University of Chemical Technology ：Natural Science, 2019, 46(6):72-77.

[7]王睿. PCB多工作板尺寸单一下料的精确算法研究[D]. 南宁：广西大学, 2019.

Wang R. Research on Accurate Algorithm of PCB Multi Working Board Single Size Cutting[D]. Nanning: Guangxi University,2019.

[8]郝玲琴. 基于神经网络的冲裁件优化排样算法研究[D]. 长春: 吉林大学,2007.

Hao L Q. Research on the Optimization Layout Algorithm of Blanking Parts Based on Neural Network[D]. Changchun: Jilin University,2007.

[9]Zheng P F, Lou J J,Lin D J, et al. Descending packing algorithm for irregular graphics based on geometric feature points[J]. Mathematical Problems in Engineering, 2020, 2020.

[10]Liu D, Teng H. An improved BLalgorithm for genetic algorithm of the orthogonal packing of rectangles[J]. European Journal of Operational Research, 1999, 112(2): 413-420.

[11]李科林. 基于临界多边形的二维不规则排样问题的研究[D]. 武汉: 华中师范大学, 2019.

Li K L. Research on Two Dimensional Irregular Packing Problem Based on Critical Polygon [D]. Wuhan: Central China Normal University, 2019.

[12]苏传生. 基于智能混合优化算法的不规则件优化排样技术研究[D]. 青岛: 青岛科技大学,2009.

Su C S. Research on Optimal Layout of Irregular Parts Based on Intelligent Hybrid Optimization Algorithm[D]. Qingdao: Qingdao University of Science and Technology,2009.

[13]李建勇, 曹月东, 鄂明成,等. 利用CHNN人工神经元网络进行排料优化计算[J]. 机械设计, 2000(5):22-24,48.

Li J Y, Cao Y D, Yue M C, et al. Using CHNN artificial neural network to optimize the layout[J]. Journal of Machine Design, 2000,(5):22-24,48.

[14]翟红岩, 苏传生,张莹,等.基于神经网络的不规则件排样技术[J].青岛科技大学学报:自然科学版, 2009, 30(4):357-360.

Zhai H Y, Su C S, Zhang Y, et al. The optimal layout technology research of irregular parts based on artificial neural networks[J]. Journal of Qingdao University of Science and Technology: Natural Science Edition, 2009, 30(4):357-360.

服务与反馈：

【文章下载】【加入收藏】