Incorporating Load Fluctuation in Feature Importance Profile Clustering for Day-Ahead Aggregated Residential Load Forecasting

Nantian Huang,Sining Wang,Guowei Cai,Jun Wang,Liang Zhang,Wenting Wang

doi:10.1109/access.2020.2971033

Nantian Huang, Sining Wang + Show 4 more

Open Access

https://doi.org/10.1109/access.2020.2971033

Copy DOI

Abstract

Residents clustering in different periods of load fluctuation and aggregated forecasting can increase the load prediction accuracy. But the strength of load fluctuation reflects the difference in the electricity consumption behavior of residents and affects the cluster results of residents. This paper presents a new day-ahead aggregated load-forecasting method for distribution networks based on the load fluctuation and feature importance (FI) profile clustering of residents. First, the input features are determined, the FI profile of residents is determined, and residents are clustered according to the FI profile. Then, the crow search algorithm is used to optimize the initial cluster centers for preventing the clustering results from falling into a local optimum. And the cluster verification index S_Dbw, the sum of the average scattering for the clusters and the inter-cluster density, is used to evaluate the cluster quality. The optimal clustering results of the aggregated load for different fluctuation periods are determined via statistical experiments. Finally, a random forest predictor based on ensemble learning is selected. According to the optimal clustering results in different fluctuation periods, a rolling forecasting model is constructed to realize day-ahead aggregated load forecasting in a residential distribution network.

Highlights

Load forecasting plays a vital role in power system, including safe operation, economic optimization scheduling, and clean energy consumption [1]–[3]
PREDICTION MODEL COMPARISON We study the application of time segmentation in clustering according to feature importance (FI) based on smart meter (SM) data at the household level for enhancing the load-forecasting performance at the distribution-network level
All the models are implemented in MATLAB running on an Intel Core i7 at 3.7 GHz with 16 GB RAM. (Analysis of the load fluctuation in the time domain and the selection method for kopt are presented in Sections II and III, respectively.) According to a comparison of the error evaluation indicators, the statistical values of the total load in the test set with time segmentation are obtained (TABLE 3)

Summary

Introduction

Load forecasting plays a vital role in power system, including safe operation, economic optimization scheduling, and clean energy consumption [1]–[3]. Accurate aggregated day-ahead load forecasting for a distribution network is important for the economic and safe operation [6] of the distribution system. Meters, and the scale of the aggregated load is generally small. The difference in residents’ electricity consumption behavior will increase the fluctuation and complexity of the aggregated load, which will have a negative impact on power quality [7]. What’s more, it will make the aggregate residential load forecasting more difficult [8], [9]

Methods

Results

Conclusion