A variance change point estimation method based on intelligent ensemble model for quality fluctuation analysis

Abstract

For multivariable production process, knowing the first time of process really changes (change point) will help to accelerate the location of assignable causes and make measures for process adjustment. So effective estimating the change point is an important way to analyse the quality fluctuation of process. In the present study, an intelligent ensemble model for quality fluctuation analysis is proposed to estimate the variance change point in multivariable process. With the method, the process is decomposed based on moving window analysis, then different types of kernel functions are combined together to form the multi-kernel support vector machine model, which has combined the feature mapping capability of each basic kernel in the new feature space. The particle swarm optimisation is considered to search the optimised multi-kernel parameters. After that, each sub-characteristic is regarded as a pattern to be recognised to determine the change point by using the optimised intelligent ensemble model. Finally, a case study is conducted to evaluate the performance of proposed approach. It reveals that the method could estimate the time of variance change point in continuous production process accurately.

Keywords:

• quality control
• process monitoring
• multivariable statistics
• control charts
• variance change point
• quality fluctuation analysis

Acknowledgements

The authors are also grateful to the editor and anonymous reviewers for their useful and helpful comments to improve the quality of the paper.

Notes

No potential conflict of interest was reported by the authors.

Additional information

Funding

This work is supported by the National Natural Science Foundation of China [grant number 51275399].