基于改进人工神经网络算法的配电网差异化节能降损方法

Differential Energy-saving and Loss-reducing Method for Distribution Network Based on Improved Artificial Neural Network Algorithm

为了解决盲目选择节能降损措施导致配电网损耗偏大的问题,研究基于改进人工神经网络算法的配电网差异化节能降损方法。设置合理的运行电压为上层规划目标,建立配电网差异化节能降损规划模型,设置安全负载约束、电流越限约束以及无功补偿设备约束,作为配电网差异化节能降损规划模型的约束条件。采用自适应遗传算法改进人工神经网络算法,针对配电网不同线路,利用改进后的人工神经网络算法求解配电网差异化节能降损规划模型,获取可令配电网差异化节能降损的最佳配电网运行参数,匹配配电网差异化节能降损方案。实验结果表明,各支路的有功损耗以及无功损耗均有明显地降低,配电网各支路网损率低于1%。

In order to solve the problem that the loss of the distribution network is large due to the blind selection of energy-saving and loss-reducing measures,a differentiated energy-saving and loss-reducing method for the distribution network based on the improved artificial neural network algorithm is studied.It sets a reasonable operating voltage as the upper-level planning target,establishes a differentiated energy-saving and loss-reduction planning model for the distribution network,and sets safe load constraints,current over-limit constraints,and reactive power compensation equipment constraints as the basis of the differentiated energy-saving and loss-reduction planning model for the distribution network.The adaptive genetic algorithm is used to improve the artificial neural network algorithm.For different lines of the distribution network,the improved artificial neural network algorithm is used to solve the differentiated energy-saving and loss-reducing planning model of the distribution network,and obtain the energy-saving and loss-reducing methods that can make the distribution network differentiated.It can obtain the optimal distribution network operating parameters to match the differentiated energy-saving and loss-reducing scheme of the distribution network.The experimental results show that the active power loss and reactive power loss of each branch are significantly reduced,and the network loss rate of each branch of the distribution network is less than 1%.