Plant-wide quantitative assessment of a process industry system’s operating state based on color-spectrum

• A general data-driven framework is proposed to evaluate the operating state of a process industry system.
• For the first time, scientific data visualization and digital image process is combined to analyze multivariate time series.
• The monitor data set is displayed based on scientific data visualization.
• The plant-wide quantitative assessment of the operating state is realized using digital image process.

This paper presents a general theoretical framework to assess the operating state of a process industry system quantitatively based on meshing the theory of scientific data visualization and digital image processing. First, a series of color-spectrum, which represent the operating state of the system, is formed by mapping the monitor data set to a group of digital color images. Second, the common feature of color-spectrum, which is named benchmark-color-spectrum, is extracted as a standard of the normal state. Third, the abnormal degree can be quantified by calculating the difference of the benchmark-color-spectrum with observed color-spectrum. At last, a plant-wide operating state of the system in a period of time can be shown by plotting quantitative abnormal degree. Two case is included to illustrate the proposed method and its appropriateness. One is a general process industry system simulator named Tennessee Eastman Process. Another is an air compressor group which belongs to a real chemical plant.