Research On Identification Of Key Quality Characteristics And Parameter Prediction Of W Company Camshaft Processing

With the continuous development of information technology,artificial intelligence is gradually applied to all walks of life,and has shown strong capabilities in the fields of intelligent recommendation,automatic identification and trend prediction.In order to achieve a higher standard of modern production,the manufacturing industry is also actively exploring the application of artificial intelligence.The application of the Internet of Things can help manufacturing companies effectively record manufacturing process data,visually monitor product processing in real time,and implement intelligent quality management methods in combination with big data and cloud computing technologies to achieve timely fault diagnosis and trace the entire process of product faults to improve the Product quality level.In recent years,statistical-based machine learning theory has been widely used in the fields of data mining,pattern recognition,and natural language processing,and has shown good classification and prediction capabilities.In the context of the implementation of the Industrial Internet of Things,product quality management of manufacturing enterprises has also become the main application scenario of machine learning theory.Using industrial big data,machine learning can establish fault prediction models and build intelligent error prevention systems to help enterprises reach zero defects level.By combing the current status of the W camshaft processing process,it is found that there are mainly problems such as difficulty in tracing the source of quality problems and difficulty in predicting the quality of key processes in the manufacturing process,mainly due to the failure to fully tap the effective information in the manufacturing process data to simplify the relationship between complex quality characteristics and quality results,the key quality characteristics identification and quality prediction of parts processing are effective methods to reduce costs and increase efficiency.Based on the data-driven thinking,this paper designs a process plan for the identification of the key quality characteristics and the prediction of the processing parameters of the camshaft machining process of W company based on the existing problems.Data-driven identification of key quality characteristics is based on feature selection theory,and the irrelevant and redundant feature dimensions in the manufacturing process data are filtered to achieve the purpose of optimizing feature indicators.Based on the relevant research and the actual scene of camshaft machining,this paper proposes a key quality characteristic identification method based on NSGA?-GBDT.This method belongs to the wrapper feature selection method,which combines the multi-objective optimization algorithm and decision tree theory.The advantage is to take the multi-classifier's classification prediction error rate and feature dimension for each quality result as the minimum optimization target,and iteratively optimize the population of quality characteristics through the NSGA II algorithm to build a non-dominated optimal frontier that meets the requirements of multiple targets and select the key quality characteristics in the camshaft machining process.Aiming at the selected key quality characteristics,and according to the characteristics of the relevant key process data,this paper builds a GM-ARMA-GBDT-based combined prediction model to reduce rework costs and prevent failures through quality prediction.This model uses the prediction data of the traditional rolling GM(1,1)and ARMA models as input signals,processes the real values as output signals to train the GBDT model,and uses GBDT to fit the nonlinear characteristics between the predicted data and the real data,based on a small amount of history time series data makes accurate quality predictions.The combination of the two-part models forms a set of key quality characteristics identification and key process parameter prediction systems for the camshaft machining process of Company W,which provides a reference for the application of machine learning theory in the field of intelligent quality management.