Research On Collaborative Scheduling Method Of Job Shop Production Logistics Network With Machine Learning

Under the background of the fourth industrial revolution driven by the new generation of information technology,intelligent manufacturing characterized by the deep integration of artificial intelligence and manufacturing has become a major trend in the development of manufacturing industry.With the change of the current manufacturing mode,it is of great significance to realize the overall efficiency improvement under the collaborative operation of jop shop production logistics in the scenario of intelligent factory,and the key to realize efficient production logistics operation is to adopt a fast and effective scheduling mechanism.Therefore,this thesis takes jop shop production logistics as the research object,base on the collaborative operation mechanism of production and logistics,combined with machine learning,supernetwork theory and virus propagation model.In this thesis,the establishment and analysis of jop shop production logistics model,job shop scheduling solution under collaborative operation of production logistics and the scheduling optimization adjustment based on the analysis of production task delay propagation are studied.Firstly,the thesis describes and analyzes the coupling characteristics of workpiece and task,task and resource,and resource and resource in job shop production logistics.Based on the relationship between various elements of job shop production logistics,the hierarchical topology model of job shop production logistics network was established from the perspective of hypernetwork theory.The topology model can effectively reveal the complex and hierarchical relationship among the multiple elements of job shop production logistics,and provide a basis for guiding the solution of job shop production logistics scheduling.Secondly,a dual BP(Back Propagation)neural network scheduling model is proposed to solve the static scheduling problem of job shop production logistics network which is coordinated between production process and logistics transportation.Based on the potential scheduling knowledge of optimal scheduling sequence data mining,the neural network scheduling model is trained.The trained dual-BP neural network scheduling model can solve the scheduling problem quickly and effectively.Then,the disturbance of the uncertainty factors in the jop shop will be directly manifested as the delay of the corresponding production task.Aiming at the process of the propagation of the production task delay on the production logistics network,the inter-layer coupling relationship of the production logistics network is introduced into the traditional virus propagation model,and a propagation dynamics model with hierarchical network coupling propagation characteristics is constructed.Further,based on the neural network scheduling results,the characteristics and range of production task delay propagation are analyzed,and the scheduling results are optimized and adjusted.On the premise of ensuring that the scheduling process runs as orderly as possible according to the original scheme,the scheduling stability can be improved by sacrificing less scheduling performance.Finally,a prototype system is developed based on the above proposed models and methods.Three modules are constructed,which are the establishment and analysis of production logistics network,the solution of neural network static scheduling and the analysis of production task delay characteristics.And the running process of the prototype system is tested through the case,and the correctness and feasibility of the model and theoretical method in the thesis are verified.