融合随机森林和神经网络的电能质量分析算法

Power Quality Analysis Algorithm Based on Random Forest and Neural Network

提出了一种融合随机森林(RF)和神经网络(NN)的电能质量分析算法.首先利用RF对电能质量信号进行特征提取和降维,然后利用NN对提取的特征进行分类和识别,最后通过实验验证了该算法的有效性,并与其他常用的电能质量分析方法进行了比较.实验结果表明,该算法具有较高的准确率、召回率和F1值,以及较快的运行速度和较低的计算复杂度.

A power quality analysis algorithm based on random forest(RF)and neural network(NN)is proposed.First,RF is used for feature extraction and dimensionality reduction of power quality signals,and then NN is used to classify and identify the extracted features.Finally,the effectiveness of the proposed algorithm is verified by experiment and compared with other commonly used power quality analysis methods.The experimental results show that the proposed algorithm has higher accuracy,recall rate and F1 value.Meanwhile,the proposed algorithm has faster running speed and lower computational complexity.