A two-level refinement method for operational rules in smart control centers

In order to improve the intelligence of power system dispatching, a two-level refinement method for operational rules in smart control centers for transmission corridors is proposed. The method includes three stages: data generation stage, in which a theme definition method, involving candidate attribute set, sample space and simulation mode, is proposed; feature selection stage, in which a heuristic regress feature selection (HRFS) method, which is suitable for continuous attributes, is introduced; rule extraction stage, in which a two-level refinement method based on linear regression tree is proposed for knowledge representation. Operational rules are built in a 4-generator 2-area test system and a realistic provincial power system using the method proposed in this paper. As the important features are selected and section rules are generated in operation sub-space, the new fine operational rules are of higher accuracy and better applicability. With the support of the operational rules, the transmission corridor capacity can be utilized more efficiently, and the control sensitivity contained in the rules is more helpful for decision-making of control strategy. © 2010 State Grid Electric Power Research Institute Press.