A New Particle Swarm Optimization Algorithm for Outlier Detection: Industrial Data Clustering in Wire Arc Additive Manufacturing

In this paper, a novel outlier detection method is proposed for industrial data analysis based on the fuzzy C-means (FCM) algorithm. An adaptive switching randomly perturbed particle swarm optimization algorithm (ASRPPSO) put forward to optimize initial cluster centroids of FCM The superiority ASRPPSO demonstrated over five existing PSO algorithms series benchmark functions. To illustrate its application potential, ASRPPSO-based exploited in problem analyzing real-world collected from wire arc additive manufacturing pilot line Sweden. Experimental results demonstrate that outperforms standard detecting outliers data. Note Practitioners —Electric (which governed by current and voltage) plays significant role monitoring operating status (WAAM) process. nominal periodic voltage may occasionally change abruptly due anomalies (such as instability, unstable metal transfer, geometrical deviations, surface contaminations), which would affect quality fabricated component. This paper focuses possible during WAAM A clustering-based anomaly where abnormal normal instances are categorized into two separate clusters. new centroid so improve accuracy. applied instances. effectiveness method. can be other applications including electrical engineering, mechanical engineering medical engineering. future, we aim develop an online system real-time defect prediction.