| Summary: | 碩士 === 國立成功大學 === 土木工程學系專班 === 93 === The power line system is prone to be severely damaged by major nature disasters such as typhoon and earthquake. These impaired power lines have to be repaired as soon as possible. Because of limited resource available, it is important to determine the repairing priority of each the damaged power line incident and the resource assigned to fixing it. Such a decision problem can be formulated as the assignment problem. In the past, mathematical programming techniques and heuristic methods are applied to modeling the repairing assignment problem. However, mathematical models suffer the computational inefficiency; on the other hand, heuristic methods are problem dependent and cannot find the accurate solution for sure. Presently, Taiwan Power Company (TPC) processes the operation of repairing assignment based on the experience of the assignment engineer without any priority rules, which could be not efficient if too many power lines are damaged. Therefore, it is necessary to establish a decision support model to streamline the operation of repairing assignment.
This research developed a genetic algorithm model to solve the proposed repairing assignment problem by maximizing the power rehabilitation rate. In addition, the VB.net language is applied to integrate the above GA-based model with the damaged reporting system which has been used by TPC. This integrated information system provides quick and reliable tool to process the repairing assignment operation. In addition, a mobile information system which can download the information of damaged power line locations and transfers TPC power line coordinate to geographic coordinate is built on the PDA. Repairing workers can use this mobile information, GPS and GIS to reach the damaged power line locations as soon as possible. Results showed that this proposed model along with the integrated information system can provide efficient solutions to the repairing assignment operations.
|
|---|
