Research On Health Assessment Technology Of Heat Treatment Equipment Based On Digital Twins

With the development of computer network technology,communication technology,and big data technology,Chinas industry is transforming towards informatization and intelligence.At the same time,China is also moving from a big manufacturing country to a strong manufacturing country.Heat treatment equipment is an important part of the industrial production line.The performance of the equipment directly affects the level of heat treatment process and thus the quality of the product will be influenced.It is of great significance to evaluate the performance of heat treatment equipment.This paper takes heat treatment equipment as the research object and proposes a digital twin-based heat treatment equipment health assessment method by combining digital twin technology with prognositics and health management technology.The main research contents are as follows:(1)The digital twin analysis framework and virtual model of heat treatment equipment are constructed.The working principle,structural characteristics and common faults of heat treatment equipment are sorted out.Based on these,a heat treatment equipment analysis framework based on digital twins is proposed,including basic support layer,data management layer,model management layer and functional application layer.Based on this framework,the twin body of the heat treatment equipment is modeled and simulated,and the accuracy of the constructed virtual model is verified through the comparison between the measurement data and the simulation data.(2)A method for monitoring the condition of heat treatment equipment based on digital twins is proposed.Based on the built digital twin of the heat treatment equipment,the simulation node is set according to the temperature measurement method of the heat treatment equipment,and the simulation data is processed in combination with the stacked autoencoder network,and the measured data and the simulation data are merged to obtain the twin data set.Based on this,a heat treatment equipment condition monitoring model based on the transferable convolutional neural network is proposed.By studying the influence of different tuning strategies on the performance of the model,the accuracy of the equipment condition monitoring is improved,and the performance of the heat treatment equipment condition monitoring method is verified through case analysis.Taking the measured data and twin data as examples,the effectiveness of digital twinning technology in condition monitoring of heat treatment equipment is verified.(3)A method for evaluating the health status of heat treatment equipment based on integrated autoencoders and self-organizing mapping network is proposed.Aiming at the problem of inaccurate equipment health assessment caused by the difficulty in extracting equipment degradation trends and the differences between the degradation modes,combined with the advantages of integrated learning and deep learning,the integrated feature extraction of heat treatment equipment data is carried out according to the feature evaluation criteria.Select the characteristics of the best class,and use the self-organizing mapping network to construct health indicators reflecting the status of the equipment to realize the health status assessment of the heat treatment equipment.The effectiveness of the proposed health assessment method is verified through case analysis,and the application effect of the digital twin technology in equipment health assessment is verified by comparing health assessment results between the actual measurement data and the twin data.