Research On Multi-objective Flexible Manu-Facturing Optimization Based On Complex Network

Under the background of the era of intelligent manufacturing,the speed of product renewal is accelerated,the production cycle is shortened,and the diversity of demand is increased.China's manufacturing enterprises are facing the problems of how to greatly improve the production efficiency,shorten the production cycle,ensure the product quality,reduce the production cost and obtain better economic benefits.In the past,our country was a manufacturing enterprise with process and discrete production modes.If we want to optimize its scheduling,we need to use various algorithms to optimize it.The popularization of big data,the application of data receivers and the enhancement of computer technology provide a large amount of available data and digital management mode for industrial production,which forces the production mode of China's manufacturing industry to change from discrete,process-oriented to data-driven.The specific work carried out in this paper mainly includes the following two points:First,in order to realize the transformation from process-oriented industry to data-driven industry,this paper uses the data of multi-objective flexible job shop production process,combines the complex network theory,and uses the complex network modeling method to establish a dynamic complex network model of multi-objective job shop based on data information.The model takes data information as nodes,the relationship between data and data as edges,and the strength of the relationship as weights of edges.It introduces dynamic changes in time and takes into account the dynamic evolution of internal nodes,edges and weights at different times.The built complex network model reflects the actual production situation of multi-objective flexible job shop from the perspective of data.The network characteristic parameters are used to prove that the model has scale-free network characteristics.The practicability of the model is verified by an example of glass fiber production in alkali-free tank furnace drawing process.After analysis,it is found that there are key bottleneck nodes that restrict production.These key bottleneck nodes are particularly important to many problems such as production cost,efficiency,time,etc.Secondly,in order to find the key bottleneck nodes in the complex network model of multi-objective flexible job shop,this paper combines the fuzzy analysis method of fuzzy multi-attribute decision-making with the network hierarchy method to obtain the fuzzy network comprehensive evaluation method.This method is used to evaluate and calculate various attributes of data nodes in complex network model,and the weight of various attributes of data nodes in process industry is obtained.Then the weights are combined with the actual attribute values of the nodes to obtain the key bottleneck nodes with the highest ranking.Finally,an example is used to verify that the method can accurately find the key bottleneck nodes,and can be better applied to multi-objective flexible job shop,which lays a foundation for realizing multi-objective flexible manufacturing optimization in the next step.