复杂零件加工过程质量控制与改进方法研究

摘要

针对航空复杂零件加工过程误差监控、溯源及过程调整问题，提出了一种复杂零件加工过程质量控制与改进方法。在传统SPC质量监控方法的基础上，设计了扩展误差流与MSPC集成的监控策略，并分别设计了单元、多元以及非线性profile的三种控制图；针对因果关系明确的几何定位误差的分解与溯源，建立了基于动态Bayesian网络的因果模型，针对误差源因果关系不明确的受力变形误差的分解与溯源，建立了基于粒子群（PSO）改进的支持向量机的误差溯源方法；在加工过程质量波动调整方面，引入多元工程过程控制（MEPC）理论，建立了MSPC与MEPC集成的加工过程波动调整方法。%Aiming at the problem of aeronautical complex part machining process error monitoring, sourcing and process adjustment, the method of complex part machining process quality control and improvement is proposed. The integrated monitoring strategy of the expanded SoV and MSPC is designed based on the traditional MSPC quality monitoring methods and the univariate, multivariate and nonlinear profile control charts are respectively designed to the three quality characteristic situation: independent, correlated, and existing function relationship. The causal model based on dynamic Bayesian network is built to the error decomposition and tracing of the geometry location error with the clear causal relation. The improved support vector machine (SVM) error tracing method based on the particle swarm optimization (PSO) is built to the error decomposition and tracing of the forced deformation with the unclear causal relation. In the aspect of machining process quality fluctuation, the adjustment method of machining process fluctuation by integrating MSPC and MEPC is proposed.