基于分群算法和人工神经元网络的配电网线损计算

本文提出了基于分群算法和人工神经元网络的计算配电网线损的实用方法。对有代表性的配电线路的线损与特征参数(如线路某时段内通过的有功功率和无功功率供电量)的样本数据,先用分群算法将样本数据分解,再用误差反向传播模型(Error Back-PropagationModel—BP模型)来映射(拟合)各个群的样本数据。考虑到BP模型固有的特性以及线损和特征参数间存在的关系,本文提出的分群算法简单,比现有方法准确,并可使其学习精度大大提高。

The Calculation of Energy Losses in Distribution Systems Based upon a Clustering Algorithm and an Artificial Neural Network Model

A new method to calculate the energy losses in distribution systems, which is based upon a clustering algorithm and an artificial neural network model, is presented. The statistical data or samples of the fenergy losses and feature factors (such as the active energy sjupply and the reactive energy supply in a period by a distribution line) for some representative lines can be used to explore the functional relationship between the energy losses and the feature factors, and the error Back-Propagation neural network model (BP model) is used to fulfil this task. In order to enhance the training (learning) accuracy of BP model, a problem-specific algorithm is proposed to divide the samples into several clusters. Simulations on an actual distribution system have shown that very accurate results can be obtained by the proposed method.