Skip to Main Content
This paper presents a control system performance evaluation method based on rough set, which uses historical data as the original decision rule data (training data). Through data discretization, decision rules merge similar terms, condition attribute simplification, calculating the core value of each of the rules, and uses the logic reasoning to achieve the training of historical data. It uses the obtained rules to achieve the training of historical data. Finally, it uses the thermal power plant operating data to proof the proposed method that it is feasible.