Product quality management based on CNC machine fault prognostics and diagnosis

International Conference on Mechanical Engineering, Automation and Control Systems 2017

机械制造;无线电电子学;计算机科学

Kozlov, A.M.^1 ; Al-Jonid, Kh. M.^1 ; Kozlov, A.A.^1 ; Antar, Sh. D.^1

Lipetsk State Technical University, 30 Moskovskaya St., Lipetsk

398055, Russia^1

关键词: Critical threshold;    Development stages;    Dynamic Bayesian networks;    Fault classification;    Integrated approach;    Intelligent fault diagnosis;    Neuro-fuzzy network;    Simulation environment;   

Others  :  https://iopscience.iop.org/article/10.1088/1757-899X/327/2/022067/pdf
DOI  :  10.1088/1757-899X/327/2/022067

学科分类：计算机科学（综合）

来源: IOP

【 摘 要 】

This paper presents a new fault classification model and an integrated approach to fault diagnosis which involves the combination of ideas of Neuro-fuzzy Networks (NF), Dynamic Bayesian Networks (DBN) and Particle Filtering (PF) algorithm on a single platform. In the new model, faults are categorized in two aspects, namely first and second degree faults. First degree faults are instantaneous in nature, and second degree faults are evolutional and appear as a developing phenomenon which starts from the initial stage, goes through the development stage and finally ends at the mature stage. These categories of faults have a lifetime which is inversely proportional to a machine tool's life according to the modified version of Taylor's equation. For fault diagnosis, this framework consists of two phases: the first one is focusing on fault prognosis, which is done online, and the second one is concerned with fault diagnosis which depends on both off-line and on-line modules. In the first phase, a neuro-fuzzy predictor is used to take a decision on whether to embark Conditional Based Maintenance (CBM) or fault diagnosis based on the severity of a fault. The second phase only comes into action when an evolving fault goes beyond a critical threshold limit called a CBM limit for a command to be issued for fault diagnosis. During this phase, DBN and PF techniques are used as an intelligent fault diagnosis system to determine the severity, time and location of the fault. The feasibility of this approach was tested in a simulation environment using the CNC machine as a case study and the results were studied and analyzed.