冲裁件优化排样算法研究  

Study on Optimal Layout Algorithm of Blanking Parts

作　　者：李方娟[1] 陈玉全[1] 李金宇 安爱民[1] 汪欢[1] 

机构地区：[1]哈尔滨理工大学机械动力工程学院,黑龙江哈尔滨150080 [2]哈尔滨恒泰汽车技术有限责任公司,黑龙江哈尔滨150060

出　　处：《哈尔滨理工大学学报》2007年第6期100-103,共4页Journal of Harbin University of Science and Technology

基　　金：黑龙江省科技攻关预研项目子课题(GC07A504)

摘　　要：对于冲裁件优化排样这个建立在多约束条件下的复杂优化问题,用单一的任何一种算法都无法找到最优解,本文提出一种用动态定位法和Hopfield人工神经网络相结合的算法,该算法在动态定位的基础上用神经网络进行进一步优化计算.实例分析结果表明,该方法与传统算法相比,不但具有自动化程度高、求解速度快等特点,而且提高了材料利用率.The optimal layout of stamping blanks, which is complicated under the control of multi-circumstances, cannot be best solved by any single algorithm, to solve this problem, based on the analysis of the traditional algorithm, a new kind of algorithm combined the dynamic positioning technique and the Hopfield artificial neural network is proposed in this paper, and with the dynamic positioning technique, the algorithm proposed in this paper goes to future optimized calculation by using the neural network. The test results show that the technique not only has the advantages of high automation and fast-footed solution, but also improves the utilization of materials.

关 键 词：神经网络 动态定位 冲裁件 优化 排样 

分 类 号：TG386.2[金属学及工艺—金属压力加工]