Research On Prediction And Processing Method Of Workshop Production Abnormality Based On LSTM

In recent years,the new generation of information technologies such as cloud computing,big data,the Internet of Things,artificial intelligence,and the mobile Internet have developed rapidly and gradually integrated deeply with manufacturing technologies.This has ushered in major opportunities for the transformation and upgrading of the manufacturing industry.Promoting the deep integration of informatization and industrialization,and then realizing the intelligentization of the manufacturing process,has become an important development strategy for China's manufacturing industry today.To this end,China has issued a "Made in China 2025" manufacturing development strategy.The integration of information technology and manufacturing technology,such as "Smart Things in the Cloud," is gradually pushing the traditional manufacturing execution system to a cloud manufacturing intelligent execution system.However,there are still some issues and challenges that need to be addressed in the process.Therefore,this article conducts in-depth research on cloud manufacturing intelligent execution system from the following three aspects.First of all,in order to realize the network and service of the manufacturing execution system,this paper proposes a cloud service-oriented method for the manufacturing execution system in the cloud manufacturing environment.This method implements cloud service for core function modules in the manufacturing execution system through the microservice package specification and the Spring Cloud technology architecture,that is,a series of cloud service operations such as registration,release,and call.In order to support manufacturing users to meet their needs through service composition.This method improves the efficiency of accessing functional modules of a manufacturing execution system and the efficiency of data transfer between modules.Secondly,in order to guarantee the quality assurance of the products manufactured by manufacturing enterprises,this paper proposes a method for predicting abnormal product quality based on deep neural networks.This method proposes the definition of product quality abnormality and evaluation criteria of product quality abnormality in discrete workshops,and analyzes the factors affecting product quality abnormality from the five aspects of personnel,materials,equipment,environment,and process flow.By establishing a BILSTM neural network based on time-series data,the abnormal factors of product quality are used as the input data of the neural network,and the abnormal values of product quality are used as the output of the neural network to predict whether the final product quality meets the quality standards during the production and processing of the product.In order to dynamically adjust the production and processing process,improve product qualification rate,and protect economic benefits of enterprises.Finally,in order to enable the production management staff in the workshop to take a scientific and comprehensive dynamic adjustment plan in a timely manner after the occurrence of an abnormal product quality,this paper proposes a method for processing abnormal product quality based on the domain knowledge graph.This method adopts a top-down design idea to analyze the discrete workshop data,and then constructs the pattern layer and data layer of the knowledge graph in turn.After the knowledge graph is established,the similarity analysis of the influencing factors of new product quality anomalies and the influencing factors of historical product quality anomalies is performed to recommend solutions to product quality anomalies,thereby achieving scientific and comprehensive discrete workshop exception handling To ensure that the workshop resumes normal production in a timely manner.To sum up,this paper adopts micro-service architecture,BILSTM neural network,knowledge graph and other new technologies,from cloud service of manufacturing execution system,intelligent prediction of abnormal product quality,and accurate processing of abnormal product quality.This paper improves the existing manufacturing execution system and gives specific solutions,which promotes the deep integration of informatization and manufacturing.