Advanced Process Control of Distributed Parameter Plants by Integration First Principle Modeling and Case-Based Reasoning : Part 2: Case-Based Reasoning Control of DPP

摘要

The Case-Based Reasoning (CBR) as a tight integration of the model-based and data-driven approaches in optimization and control of plants with distributed parameters (DPP) is considered into this second part of the paper. The main pecularities of the adopted modifications of CBR in accordance of the area of application are examined – structuring of search space as virtual version space, retrieval for different form of attribute-value presentations, efficient indexing and adaptation using local regression in order to improve retrieved solution into the direction to obtain query problem solving. The main results of the accomplished simulations are reported focusing both the maximum capacity and minimum heat consumption. The results show that suboptimal control of TTP based on modified CBR is a reasonable approach.