Integration of renewable energy and new load (such as electric vehicle) has a great influence on the grid, and traditional load model is not appropriate for analysis, therefore a method for generalized load modeling is proposed. The procedures are segmented into two parts: generalized model clustering and neural network load modeling. Based on the generalized load data collected, K-means clustering is used for studying the time-varying characteristics of generalized load and power consumption performances of users. Then after clustering, the corresponding data serve as the training set and validation data for generalized load modeling with Radial Basis Function (RBF) neural network. The method is proved effective and accurate on a certain area with hourly generalized load data in a year.
Paper
Full text
Generalized Load Modeling Method Based on Clustering and Neural Network
Semantic Scholar · Engineering · 2020
Abstract
Integration of renewable energy and new load (such as electric vehicle) has a great influence on the grid, and traditional load model is not appropriate for analysis, therefore a method for generalized load modeling is proposed. The procedures are segmented into two parts: generalized model clustering and neural network load modeling. Based on the generalized load data collected, K-means clustering is used for studying the time-varying characteristics of generalized load and power consumption performances of users. Then after clustering, the corresponding data serve as the training set and validation data for generalized load modeling with Radial Basis Function (RBF) neural network. The method is proved effective and accurate on a certain area with hourly generalized load data in a year.