基于GIS空间查询和遗传算法的变电站规划

Substation Planning Based on GIS Spatial Query and Genetic Algorithm

针对遗传算法应用于变电站规划时产生大量不可行解的缺点,提出了一种地理信息系统(GIS)空间查询和遗传算法相结合的新算法。该算法设计了一种能够保证全局最优解的自适应交叉和变异概率,利用地理对象的拓扑关系进行空间查询,特别是在不可行解的处理中用GIS提供的空间查询功能来判断新建变电站是否在规划区域内且在适宜新建区域内,并将查询结果反映到适应度函数的惩罚函数中。经算例验证,此方法能够有效地减少不可行解,最终收敛到全局最优解,使得规划结果符合实际要求。

Considering the genetic algorithm（GA） producing many infeasible solutions in substation planning,a new algorithm with the combination of geographic information system（GIS） spatial query and GA was proposed.An adaptive probability of crossover and mutation was designed in the algorithm to ensure the global optimal solutions.The topological relationship of geographic objects was used to conduct the spatial query.Particularly,the spatial query function provided by the GIS was used in solving infeasible solutions to judge whether the new substations were in the planned and suitable area,and the query results were reflected in the penalty function of fitness function.By the case verification,it was showed that this method could effectively reduce the infeasible solutions and finally converge to the global optimal solutions,which made the planning results satisfy the practical requirements.