制造装备智能优选及快速的设计方法

Research on Intelligent Optimization and Rapid Design Method of Manufacturing Equipment

针对企业在制造装备选择效率低、现有数据利用不足的问题。本文提出了一种改进的自组织特征映射(SOM)神经网络联合模糊近似优先比的方法。基于数据挖掘技术,对制造装备优选数据预处理方法开展了研究,改进了SOM神经网络中欧氏距离计算方法,引入贡献率对历史方案数据进行加权处理,构建了多特征数据的融合模型,联合了模糊近似优先比方法对多方案进行相似性度量,获得了最优设备参数信息,通过二次开发技术实现了产品的快速设计。以制麦生产线输送设备为例,验证了该方法的可行性和有效性。

Aiming at the problems of low efficiency in equipment selection and insufficient utilization of existing data,a method of an improved self-organizing feature mapping(SOM)neural network combined with fuzzy approximate precedence ratio is proposed in this paperBased on data mining technology,the research is carried out on the preprocessing method of manufacturing equipment optimization dataThe distance calculation method in the SOM neural network is improved,and the contribution rate is introduced to weight the historical program dataA multi-feature data fusion model is established,and the fuzzy approximate priority ratio method is combined to measure the similarity of multiple programs to obtainthe best parameterA rapid product design is achieved through the secondary development technologyThe feasibility and effectiveness of the method are verified through the example of the conveying equipment of the malting production line.