Development of Simulator with Artificial Intelligence on Secondary Power System Operation

Abstract

Centralized control from a control center is popular in secondary power systems. However, as operating systems are modernized and automated, operators rarely have opportunities to experience various situations. On the other hand, operators are expected to quickly cope with any situation once an accident happens. In this paper, artificial intelligence has been applied to the training of a secondary power system operation. The developed system has simulation, training, and inquiry modes. In the training mode, it helps operators train in restoring power systems. If an operator succeeds in restoration, the system evaluates executed operations. If he fails, the system generates and presents an example of restoration operations. These functions have been realized by expert systems of a production rule type. Methods of automatically generating restoration process and of evaluating executed operations have been described in detail. Restoration operations are generated so as to restore blackout load buses one by one. The used rules are very simple, and are easy for operators to understand. On an operator's request, it explains how each operation is generated. Evaluation is made on items such as new blackouts and overloads caused by executed operations. If any operations come under these items, then the system gives some warnings and corrections about them. The system was applied to an example power system, and its effectiveness was verified.