基于混沌粒子群算法的变压器经济运行方式的研究

Study of Economic Operation for Transformer Based onChaotic Particle Swarm Algorithm

为寻求多台并联变压器的最优经济运行方式,在考虑变压器负荷越限和绕组最热点温度限制约束条件基础上,提出了一种利用混沌粒子群算法寻找最优经济运行方式的策略。该算法在混沌粒子群优化算法中引入交叉变异过程,有效地处理了变压器投切时间的约束并提高了粒子群优化算法的全局收敛能力。以某110kV变电站为例,与其他控制策略相比,利用混沌粒子群算法进行寻优求解时,既保证了单台变压器负荷不越限及绕组绝缘不受损,又达到了有功功率损耗最小的目的;策略所用到的优化算法易于实现;利用算法求解过程中所引入的交叉和变异操作有效避免了寻优结果陷入局部最优解的问题。

In order to seek the optimal economic operation strategy of multiple parallel transformers,a method using chaotic particle swarm optimization algorithm based on the consideration of the transformer load limit and the winding hotspot temperature limit constraints is proposed.The crossover mutation process is introduced into the chaotic particle swarm optimization,which can effectively deal with the transformer switching time constraint and improve the global convergence ability of particle swarm optimization.Taking a 110 kV substation as an example,compared with other control strategies,when using the chaotic particle swarm optimization algorithm for optimization,it does not only ensure that the load of a single transformer restricted within the limit and the winding insulation is not damaged,but also achieves the purpose of minimizing active power loss.The optimization algorithm used in the strategy is easy to implement.Moreover,the crossover and mutation operations introduced effectively avoids the problem of the optimization result falling into the local optimal solution.