基于四维可视化算法的牵引变电所位置优化

针对牵引供电系统设计的复杂性以及传统优化设计方法过程繁琐且很难找到可靠的最优解,粒子群和遗传算法等的迭代过程复杂并且容易陷入局部最优解的问题,将一种四维可视化算法运用到牵引供电系统的优化设计中。四维可视化算法不需要设定初始解、能够同时展现全局最优解的分布、能判断解的稳定性、方便添加约束条件。四维可视化优化算法通过编程使目标函数的最优解集在四维数据场可视化。以接触网对地电压为约束条件,以牵引网有功功率损失最小为目标函数,运用四维可视化算法进行牵引变电所和分区所位置的优化设计。通过实例计算验证了四维可视化算法可有效降低牵引网的功率损失。研究表明该方法能很好地克服传统方法和粒子群等优化算法在牵引供电系统优化设计中的不足。

Location Optimization of Traction Substation Based on Four-dimensional Visualization Algorithm

Because of the complexity of the design of traction power supply system and the tedious process of traditional optimization design method, it is difficult to find reliable optimal solution. The iterative processes of particle swarm optimization and genetic algorithm are complex and likely to fall into local optimal solution. A four-dimensional visualization algorithm is applied to the optimal design of traction power supply system. The algorithm does not need to set the initial solution. It can simultaneously show the distributions of the global optimal solutions and judge the stability of the solution, and it is easy to add constraints. The optimal solution set of the objective function is visualized in the four-dimensional data field by programming the four-dimensional visualization optimization algorithm. Taking the catenary to the ground voltage as the constraint condition and the minimum active power loss of the catenary as the objective function, this paper uses the four-dimensional visualization algorithm to optimize the position of traction substation and sectioning post. The calculation of examples verifies the efficiency of the four-dimensional visualization algorithm in reducing power loss of the catenary. The research results show that this method can overcome the shortcomings of traditional methods and particle swarm optimization algorithms in the optimal design of traction power supply system.