Research On Intelligent Quality Control Method Of Multi-variety And Small-batch Manufacturing Process

Due to the production characteristics of multiple varieties,single-piece and small batch mixing,and multiple types cross-parallel of aerospace complex components,the problems such as multiple varieties of cross-parallel,single-piece and small batch mixing,intangibility distribution of the quality characteristic data,insufficient sample data and insufficient quality control effect in manufacturing process are facing in the quality control of machining process.With the rapid development of China’s aerospace industry and the transformation and upgrading of aerospace enterprises,the problems become particularly prominent.Therefore,it is crucial to improve product quality consistency and qualified rate in manufacturing process,reduce costs and ensure quality for aerospace manufacturing enterprises by exploring the product quality control methods for multi-variety and small-batch manufacturing process.Thus,this thesis takes an aerospace machinery processing company as the research background,and then studies the quality control method of key process in manufacturing process of typical Aerospace complex components such as complex structure frame,thin-walled plate and cabin.The specific research contents are as follows:(1)To solve the problems of hardly identification of the key processed caused by multi-variety cross parallel,single-piece and small batch mixing,dynamic process and various processes in the aerospace complex component manufacturing process,a quality key process identification model based on clear comprehensive evaluation and grey relational analysis is presented.To handle the problem of insufficient sample size of key processes,the key process clustering analysis index system based on 5M1 E is established to realize the hierarchical clustering analysis of key processes and preliminarily expand the quality data sample set.And then lay the foundation for subsequent quality control and diagnosis.(2)To cope with the problems of uncertain quality data distribution,insufficient data samples and difficult detection of process drift in the aerospace complex component manufacturing process,A non-parametric adaptive dynamic EWMA control chart multi-objective intelligent optimization design method based on improved artificial fish swarm algorithm is proposed.First,a dynamic sampling method according to clustering distance of key processes is designed,and then a non-parametric adaptive dynamic EWMA control chart is constructed based on U statistics.Moreover,the Markov chain model is introduced to analyze and calculate the statistical and economic evaluation indexes of the control chart.And the NAD-EWAMA control chart parameters optimization model based on the cloud clear comprehensive evaluation method is formulated to solve the problems of slow convergence speed and accuracy of optimization results in the later stage of the algorithm.In this way,the optimal control chart parameters combination is obtained,and the effective monitoring of the quality of the key process manufacturing process is achieved.(3)To handle the problem that many and fine quality anomalies in the manufacturing process of aerospace complex components,a multi-variety and small batch quality intelligent diagnosis method based on key process quality gene and cloud CBR technology is proposed.The hybrid model of time series is used to improve the precision of pattern feature of control chart and the construct the quality gene of key process.As a result,the parametric quantitative description of manufacturing process quality is realized.Furthermore,the learning knowledge of historical cases and new cases are integrated by analyzing the similarity of quality genes in key processes.Finally,a multi-variety and small batch quality anomaly diagnosis model is constructed based on the cloud clear comprehensive evaluation and CBR technology,then the traceability and adjustment of quality intelligent anomalies is achieved.