The Real-time Big Data Processing Method Based on LSTM for the Intelligent Workshop Production Process

摘要

With the wide application of intelligent sensors and the Internet of things in the intelligent workshop, a large number of real-time production data are collected. Accurate analysis of the collected data can help producers to make effective decisions. Compared with the traditional data processing methods, artificial intelligence, the main method of big data analysis, has been applied in the manufacturing industry increasingly. However, it is different for Artificial Intelligence models to process real-time data from intelligent workshop production. Thus, this paper proposes a real-time big data processing method based on long short term memory (LSTM) for the workshop production process. This method takes the historical production data extracted by the Internet of things workshop as the original data set, preprocesses the data, and uses the LSTM model to train and predict the real-time data of the workshop. Finally, the experimental results are compared with K-NearestNeighbor (KNN), Decision Tree (DT) and traditional neural network model. The results show that the prediction accuracy of the LSTM model is 126.9%, 21.1%, and 14.7% higher than that of the traditional neural network, KNN, and DT respectively.