A hyper-heuristic inspired approach for automatic failure prediction in the context of industry 4.0. (September 2022)

• Record Type: Journal Article
• Authors: Navajas-Guerrero, Adriana; Manjarres, Diana; Portillo, Eva; Landa-Torres, Itziar
• Is Part Of: Computers & industrial engineering. Volume 171(2022)
• Publication Date: 2022-09
• Subjects: Collective Anomaly detection -- Time series analysis -- Multiple parameter optimization -- Cold stamping process -- Fault detection -- Fault prediction; Engineering -- Data processing -- Periodicals; Industrial engineering -- Periodicals; 620.00285
• DOI: 10.1016/j.cie.2022.108381
• Languages: English
• ISSNs: 0360-8352

Abstract

Abstract: In the era of technological advances and Industry 4.0, massive data collection and analysis is a common approach followed by many industries and companies worldwide. One of the most important uses of data mining and Machine Learning techniques is to predict possible breaks or failures in industrial processes or machinery. This research designs and develops a hyper-heuristic inspired methodology to autonomously identify significant parameters of the time series that characterize the behaviour of relevant process variables enabling the prediction of failures. The proposed hyper-heuristic inspired approach is based on the combination of an optimization process performed by a meta-heuristic algorithm (Harmony Search) and feature based statistical methods for anomaly detection. It demonstrates its adaptability to different failure cases without expert domain knowledge and the capability of autonomously identifying most relevant parameters of the time series to detect the abnormal behaviour prior to the final failure. The proposed solution is validated against a real database of a cold stamping process yielding satisfactory results respect to a novel A U C _ R O C based metric, named A U C _ M O D, and other conventional metrics, i.e., Specificity, Sensitivity and False Positive Rate. Highlights: A new Hyper-heuristic based on machine learning and feature based statistical methods. Optimizing failure prediction parameters: window size, TS features and thresholds. Abnormal patterns in the time window prior to the failure are potential fault predictors. A proposal of a new evaluation metric based on AUC-ROC with a penalization factor. The approach is validated in a real case of fault prediction in a cold stamping press.