Multi-mode plant-wide process operating performance assessment based on a novel two-level multi-block hybrid model

Abstract

Process operating performance assessment judges the optimal degree of the process performance and identifies the cause for non-optimality in real-time. Therefore, process operating performance assessment is of great significance in both the theoretical research and practical application. In this article, the multiple working modes, the plant-wide process characteristics, and the coexistence of both the quantitative and qualitative information are considered. To solve the above problems, a novel two-level multi-block hybrid model based operating performance assessment approach is proposed. Under each working mode, the two-level hybrid model is established. Due to the novel offline modeling method, the global performance grade is directly determined by the sub-block performance grades and the sub-blocks can be modeled separately with different modelling techniques. In this article, a quantitative information dominated sub-block is modeled by the quantitative method, while a qualitative information dominated sub-block is modeled by the qualitative approach. When any non-optimal operating performance occurs, a contribution rate based non-optimal cause identification technique is developed within the non-optimal sub-blocks. At the end, the developed assessment approach is applied to a gold hydrometallurgy process to illustrate its feasibility and validity.