원문정보
초록
영어
Fault diagnosis for quality control during the multivariate production process is widely used to detect abnormal fluctuations, find out failure reasons and take measures to maintain the stability of the production system accordingly. The neural network methodology has become a main method in the field of intelligent diagnosis recently. However, it has certain deficiencies such as longer training time, slower convergence rate and easier falling into a local optimal solution easily. As a result, the effect of fault diagnose is influenced. Thus, this article proposes the idea of using the Fruit Fly Optimization Algorithm in the multivariable process fault diagnosis model, at the same time, to analyze the out of control sample data in the automobile crankshaft production. Compared with the neural network model in dealing with the fault diagnosis in multivariate process, Fruit Fly Optimization Algorithm’s effectiveness s verified.
목차
1. Introduction
2. Research Methodology
3. Presented Multivariate Production Process Failure Diagnosis Algorithm
4. Case Application
4.1. Applying FOA for Failure Mode Diagnosis
4.2. Applying BP Neural Network BP for Failure Mode Diagnosis
4.3 Diagnosis Results and Comparison
5. Concluding Remarks and Future Research Directions
Acknowledgements
References