Comprehensive Evaluation of Power Quality via Sequential Floating Search Method

Comprehensive assessment of power quality is the basis for measuring power quality. SFSM (Sequential Floating Search Method) with Bhattacharyya distance is used to select the most effective power quality evaluation indices, then use three classical classifiers to verify the validity of the selected indices. Through analyzing the data collected by the monitoring points, each evaluation indicator is evaluated and screened. The reserved evaluation indices are searched by SFSM algorithm with the Bhattacharyya distance. SVM (support vector machine), KNN (K-nearest neighbor) and RBF (radial basis function) neural network are used to test the classification accuracies of the selected evaluation indices. The experimental results show that this method can find the best combination of power quality indices, which can reasonably evaluate the comprehensive evaluation level of power quality for a monitoring point.