Research On The Identification Method Of Production Equipment Operation Status Based On Time Series Data Cluster Analysi

In the era of Industry 4.0,the manufacturing industry is facing a comprehensive change,and it is necessary to explore new manufacturing modes and technical means to adapt to the development trend of digitalization,networking and intelligence.Therefore,the application of data analysis and machine learning in intelligent manufacturing has important theoretical and practical significance,which can improve the efficiency and quality of manufacturing industry,promote the implementation of intelligent manufacturing strategy and deepen the enterprise application.In the era of Industry 4.0,the manufacturing industry is facing a comprehensive change,and it is necessary to explore new manufacturing modes and technical means to adapt to the development trend of digitalization,networking and intelligence.Therefore,the application of data analysis and machine learning in intelligent manufacturing has important theoretical and practical significance,which can improve the efficiency and quality of manufacturing industry,promote the implementation of intelligent manufacturing strategy and deepen the enterprise application.In order to solve the problems of low similarity accuracy and redundant data affecting computing efficiency,a method of time series clustering analysis and state recognition based on data analysis and machine learning was proposed in this paper.Specific research contents are as follows:(1)In order to identify the state of manufacturing production equipment,this paper analyzes the research in the field of time series analysis and industrial state recognition,summarizes the models and methods related to time series and clustering,and provides a theoretical basis for the subsequent cluster analysis and state recognition of production equipment time series.(2)Aiming at the similarity measurement problem of time series of production equipment,this paper proposes a clustering algorithm based on k-medoids--DTWkmedoids.This algorithm improves the flexibility of time series alignment by realizing dynamic time warping.On this basis,a threshold mechanism is built to recognize different states of production equipment.It provides an effective method for the supervision and testing of industrial production.(3)This paper proposes a new Optimized Evaluate Cluster(OEC)for low computational efficiency caused by large amounts of noise,missing values and irrelevant data in production equipment data.A new u-shapelets quality measurement method is used in the model to unsupervised discover features from time series and discover underlying data states from original time series.The number of clusters was determined by elbow method,and finally the clustering was performed by SDTW distance space,which improved the overall quality of u-shapelets collection,and the accuracy and computational efficiency of the final clustering.(4)Based on the above research results,this paper developed a time series cluster analysis and state recognition system for production equipment.Users can input relevant parameters according to requirements,and realize the functions of time series data collection,optimal subsequence extraction,similarity calculation,cluster analysis and state recognition of production equipment.