Research And Application Of Gear Hobbing Carbon Footprint Based On Association Rules Mining

As a major manufacturing country,if our country wants to reduce carbon emissions in the field of machinery manufacturing,large and medium-sized backbone enterprises across the country need to make great efforts in low-carbon manufacturing research,especially in the field of gear manufacturing.As a common transmission component,gear occupies an important position in the machinery manufacturing industry.Gear tooth profile processing methods and machine tools are diverse,and hobbing machine is used frequently.The gear hobbing process consumes a lot of materials and energy,and there are many factors affecting carbon emissions.In recent years,research on gear carbon footprint accounting has accumulated a large amount of carbon footprint data,but the research on these carbon footprint data has rarely led to ineffective use.In order to dig out the relationship behind the carbon emission factors in the gear hobbing process from these data,and provide useful information for corporate decision makers to achieve low-carbon gear hobbing manufacturing.Based on a large amount of carbon footprint data,this paper uses the commonly used correlation analysis method of data mining to study the gear hobbing process.First of all,This paper selects the carbon footprint of gear hobbing as the application object of association rule mining,uses the gear manufacturing enterprise information system as the source of carbon footprint data or basic data,and obtains carbon footprint data from PDM,CAx,MES and other components.Analysis of gear hobbing processing such as gear hobbing processing principle and gear hobbing machine transmission movement,it is clear that the main carbon emission sources in the gear hobbing process are gear hobbing machine tools,gear workpieces,hob cutters and cutting fluid.Summarize the influencing factors of carbon emission according to the gear hobbing process and the source of carbon emission,and the gear hobbing machine tools,gear workpieces,hob cutters,cutting fluid,cutting amount or other carbon emission influencing factors and the correlation between these influencing factors are analyzed.Secondly,the data obtained by carbon footprint calculation may have missing values,abnormalities such as noise and redundancy,or the form of storage in the gear enterprise information system is not uniform,and the carbon footprint data set may be too large for mining.Considering that in association rule mining,in order to accurately and efficiently dig out the potentially valuable information of the hobbing carbon footprint,the obtained data needs to undergo a data preprocessing process,including data cleaning to obtain a large data set after missing,noisy and redundant processing;data transformation allows the associated attribute value to be between specific partitions,which is convenient to convert into a format that is easy to store;data reduction achieves the purpose of data compression through reduction or dimensionality reduction,and improves the efficiency of association mining while ensuring that the association results are not affected.Next,a gear hobbing carbon footprint data warehouse was constructed on the data obtained through a series of preprocessing operations.In view of the fast response of search and the efficient principle of storage,the MySQL database was selected for storage,and the gear hobbing carbon footprint data warehouse was compared with OLAP to realize the improvement of the efficiency of mining carbon footprint association rules.Finally,use association rules to discover the potential connections among the four influencing factors of the carbon footprint characteristics of gear hobbing machine tools,gear workpieces,hob cutters and cutting fluid,and establish an carbon footprint rule model T _am(ge ar)of gear hobbing according to the whole process.Examples show that Apriori algorithm can greatly reduce the unnecessary properties in practice.Aiming at the time-consuming shortcomings of the Apriori algorithm,in order to reduce the mining time,the improved mining algorithm FP growth is adopted and proved in practice.Through the attribute or data set difference degree to quantify the error of continuous new data on the old data model.The information finally mined can provide more reasonable low-carbon manufacturing suggestions and data support for the decision-makers of gear manufacturing enterprises.After completing the above steps,a mining system based on carbon footprint data of hobbing process for gear manufacturing enterprises is developed,which allows complex mining operations to be implemented by the system,and enterprise personnel can quickly realize data mining with some simple interface operations.