用于电力数据管理分析的负荷预测与异常检测

构建了一个电力数据管理分析系统,并设计了电力负荷预测算法和异常数据检测算法问题.首先,针对BP神经网络在预测电力负荷存在的因初始权值与阈值设置影响估计精度的问题,提出利用粒子群优化BP神经网络网络参数,提高了预测算法的收敛速度与预测精度;然后,针对电力数据异常检测算法效率较低的问题,提出了基于改进谱聚类的异常数据检测算...

Load forecasting and anomaly detection for power data management system

A power data management and analysis system is constructed,and the power load forecasting algorithm and abnormal data detection algorithm are designed.Firstly,BP neural network has the problem that the initial weights and threshold settings affect the estimation accuracy in power load forecasting.In order to improve the convergence speed and the prediction accuracy of the prediction algorithm,the particle swarm optimization(PSO)is proposed to optimize the parameters of BP neural network.Secondly,the efficiency of the existing power data anomaly detection algorithms is low.An abnormal data detection algorithm based on improved spectral clustering is proposed to improve the detection efficiency of power abnormal data.Finally,the effectiveness of the proposed algorithm is verified by testing the actual power data.