The Distribution Network Planning Based On Geographic Information And Risk Assessment

Distribution network planning is designed to solve the problems existing in thecurrent distribution network, and tease the grid structure of the distribution network.Under the premise that the planning meets the needs of load growth, the planning schemeshould improve the indicators of distribution network operation. At the same time theplanning focuses on perfecting and optimizing the network frame structure, coordinatingthe subordinate power grid with the superior and enhancing the load transfer ability.Therefore, the study of distribution network planning is very necessary. In this paper, themain work includes several parts as the following:(1)Substation locating and sizing should consider the influence of geographicfactors such as land properties, traffic conditions, construction conditions. In this paper,geographic information factor model is constructed based on the improved ideal pointmethod, and the model is embedded into the construction and operation costs. To solvethe model, the initial population is divided into multiple sub populations. Amulti-population differential cultural algorithm based on chaotic knowledge migration ispresented. The algorithm designs four kinds of spatial knowledge guide evolution. Andknowledge migration interval is dynamically adjusted according to evolution time. Thenchaos sequence is used to deeply explore migrated cells for the improvement of thevalidity of migrated knowledge.(2)On the basis of processed geographic background picture, considering therandomness and fuzziness of the load, the distribution network planning model based oncredibility theory is put forward. The encoding pattern is layer by layer and line by line.A genetic algorithm based on the combination of improved fast and elitist non-dominatedsorting and multi-objective chaos weighting is presented. Partheno-Genetic Algorithm isused to guarantee the individual being radial pattern. There is no need to restore theinfeasible individuals, which increases optimization efficiency.(3)Based on the analysis of all kinds of risk source in the distribution networkplanning this paper established a risk evaluation index system of power distributionnetwork planning. A comprehensive evaluation model is constructed based on improvedvariable weight and matter-element extension model. The variable weight theory is usedto determine the weight of each rating index, reducing the subjective factors in the riskassessment and reflecting the active participation of the evaluation objects in thecomprehensive evaluation.