GIS与粒子群算法在农村变电站选址规划中的应用

Application of GIS and particle swarm optimization in rural substation locating

针对基本粒子群算法在农村变电站选址问题中得到全局最优解的收敛速度慢和易陷入局部最优解的缺点,该文以某县开发区为例,运用惯性权重动态调整策略,有效地平衡了算法的全局和局部搜索能力,从而改善了基本粒子群算法的性能,并且充分考虑地理信息系统对规划站址的影响,将改进的粒子群算法和图形问题相结合。结果表明:基本粒子群算法得到最优解的迭代次数为48次,改进后算法的迭代次数减少到26次,得到最优解的速度提高了近一倍,并以GIS为平台实现了规划的可视化。

Considering the defects of Particle Swarm Optimization （PSO） in the optimization of rural substation location, which constringency speed of getting the global optimum is slow and that is easy to fall into local optimum, this study took development zone of one county as an example to optimize the substation location. The inertia weight dynamic adjustment strategy was utilized to balance effectually the global and local search ability, which greatly improved the capability of PSO algorithm, By considering the influence of Geographic Information System （GIS） on the substation site, the improved PSO was combined with graph problem. Results show that the iteration times of PSO is 48 and the iteration times of the improved PSO reduces to 26. The speed of getting global optimum has doubled. The visualization of the planning process was realized by GIS.