【摘要】 随着我国电能需求量不断提升，配电网可靠性要求逐步提高，配电网设备状态研判难度也不断增大。针对该问题，本文提出一种基于群体优化-概率神经网络的配电网设备状态研判模型。引入改进后的人工鱼群算法对概率神经网络的平滑因子进行寻优，避免其因随机设置而导致研判精度不理想的问题。基于群体优化-概率神经网络算法建立设备状态研判模型，同时利用合成少数类过采样技术改善配电网数据集不平衡的问题，采用主成分分析法对数据集进行特征属性指标提取，减少冗余指标对状态研判精度和时间的影响。实验结果表明，本文模型在状态研判的精度和计算时间上均具有一定优势，能够在配电网的状态研判过程中起到辅助作用。更多还原

【Abstract】 With the continuous increase of electrical power demand in China, the reliability requirements of distribution network are gradually increased. Meanwhile, the research of distribution network equipment state is getting harder and harder. In order to solve this problem, this paper proposes a distribution network equipment state estimation model based on swarm optimization-probabilistic neural network. The improved artificial fish swarm algorithm is introduced to optimize the smoothing factor of probabilistic neural network to avoid the problem of acquiring unsatisfactory accuracy due to random setting of the parameters. The equipment state estimation model is established based on the population optimization-probabilistic neural network algorithm, and the synthetic minority class oversampling technology is used to improve the unbalanced problem of the distribution network data set. Principal component analysis is used to extract the characteristic attribute indexes of the data set to reduce the influence of redundant indexes on the accuracy and time cost of the model. The experimental results show that the proposed model has advantages in the accuracy and calculation time of equipment state estimation which means the model could play an auxiliary role in the process of state estimation of distribution network.更多还原