Application of GA to Determining Network Configurations for Outage Works by EOSPF+TS Method

As a power system is becoming larger and more complex, it has become more difficult for planning engineers to adjust outage works since the number of maintenance works has increased accordingly. For assisting planning engineers, the authors have developed a supporting expert system for outage work planning of electric power facilities. In this system, a suitable system configuration for a particular set of outage works is chosen out of the prepared system configurations which have been formed from the past operation records. However, in order to obtain feasible plans from the standpoint of adequacy in system configurations, it is necessary to consider the following three issues: (a) reducing the number of switching operations, (b) transition of system configurations from day to day, and (c) complete avoidance of supply shortage at N-1 contingencies. In order to achieve the above objectives, the authors have proposed an approach of generating system configurations under N-1 security constraints, which is based on the extended optimal switching power flow (EOSPF) and Tabu Search (TS) algorithm. However, since the proposed approach minimizes the objective function with respect to an N-1 security constraint violation, it is impossible to minimize switching operations for overall planning period. In this paper, we propose a method of making up system configurations by means of a combination of (EOSPF+TS) and Genetic Algorithm (GA), in which the proposed method minimizes the number of switching operations for overall period while satisfying N-1 security constraints. In order to confirm the effectiveness of the proposed method, it has been applied to a 42-bus system with successful results.