POM SPACE REDUCTION AND RECURSIVE CLUSTERING FOR MONITORING OF INDUSTRIAL PROCESSES

Authors

Abstract

Projected Orientation Map (POM) is a tool proposed to support monitoring on high dimensional industrial processes. The reduction of n-dimensional space to bidimensional one is developed through multidimensional scaling with a proposed iterative capability. The space reduction technique is aimed to keep the relevant and informative geometric characteristics of the original space and builds a 2-D map (POM), where the operational regions of the process under specific conditions can be identified. In the bidimensional space several operating regions are identified by Dignet recursive clustering method, giving information to the monitoring system about the possible quality of the running operating conditions. This method is applied to the process of Hydro Desulfuration (HDS) from Refinery of Petrogal at Sines (Galp Energia).

Keywords

supervision and monitoring; data mining; space reduction; recursive clustering; refinery.

Subject

supervision and monitoring data mining

Related Project

CLASSE - Classificação Sintética para Supervisão Industrial (Synthetic Classification for Industrial Supervision)

Conference

Controlo 2006, 7th Portuguese Conference on Automatic Control, September 2006