| Summary: | 碩士 === 國立高雄應用科技大學 === 電機工程系 === 99 === Distribution system outage scheduling is to set up the crew dispatching for the service areas that would be tripped due to engineering reasons such as repairing, replacement or system extension. It is important to have optimal scheduling for the service areas and the engineering crews not only to improve service availability but also decrease the expenditure for the works. It is inherently discrete, temporal, and multi-objective problem.
In this thesis, an immune genetic algorithm is proposed to deal with the optimal outage scheduling of distribution systems with integration of Google maps. There are mainly two types of lymphocytes, namely B cells and T cells play different roles in the immune genetic algorithm. The characteristics of specificity, memory, and adaptiveness of lymphocytes are modeled and integrated with genetic algorithm in the proposed method to effectively solve the problem. The fitness function is to minimize the engineering days, the outage loading, the difference of working time among the crews, and the distances of routings. Improved crossover rules and a weighted dynamic mutation method are presented. The transportation time and distance obtained from Google-Maps are integrated in the scheduling approach. Mobile phones applications are used in the field to communicate with the dispatching center with the scheduling displayed on the Google-Maps.
Parts of the distribution system located in Kaohsiung area is selected for the computer simulations by the proposed genetic algorithm and immune genetic algorithm. Comparison of the solution efficiencies and evolutionary trends by ordinary genetic algorithm and immune genetic algorithm are implemented. Form the simulation results, it shows the immune genetic algorithm models is not only feasible but also effective improve genetic algorithm disadvantage. Also, highly reliable scheduling results can be obtained effectively.
|
|---|
