摘要
加工中心占地是机床布局方案设计中考虑的关键因素,为使获得最小占地面积的加工中心布局方案,提出基于部件间约束的加工中心布局预测方法。将加工中心的三维模型中主要部件投影为二维矩形布局图元,以二维布局图元中矩形的型心坐标及转角作为输入,加工中心占地面积作为输出训练BP神经网络预测模型,建立适应度函数。考虑部件间、操作人员的操作空间与部件干涉问题,给出不干涉约束以及相邻约束,通过遗传算法进行求解得到加工中心最优布局方案。最后以实例验证了所提布局预测方法的有效性。
The machining center occupying area is the key factor considered in the design of machine tool layout. In order to obtain the layout plan of the machining center with the smallest area, a method of machining center layout prediction based on the constraint among components is proposed. The main components in the 3 D model of the machining center are projected as two-dimensional rectangular layout primitives, the center coordinates and corners of rectangles in the two-dimensional layout primitives are taken as input, and the processing center footprint is taken as output to train the BP neural network prediction model, and establish fitness function. Considering the inter-component, operator＇s operating space and component interference problems, this paper gives non-interference constraints and adjacent constraints, and uses genetic algorithms to solve the optimal layout of the machining center program. Finally, an example is given to verify the validity of the proposed layout prediction method.
作者
李纬天
萨日娜
LI Weitian;SA Rina(College of Mechanical Engineering, Inner Mongolia University of Technology, Hohhot 010051, China)
出处
《机械工程师》
2018年第5期28-32,共5页
Mechanical Engineer
基金
国家自然科学基金(51405247
51765052)
内蒙古自然科学基金(2016MS0507)
关键词
机床布局
矩形图元
BP神经网络
遗传算法
machine tool layout
rectangle graph unit
BP neural network
genetic algorithm
