Line Loss

Line loss, the energy dissipated during electricity transmission and distribution, is a significant concern requiring effective management strategies. Current research focuses on improving the accuracy of short-term line loss rate forecasting, employing advanced techniques like attention mechanisms integrated with graph convolutional networks and long short-term memory networks to capture complex spatial and temporal dependencies within power grids. These improved forecasting models aim to optimize resource allocation and enhance the efficiency of power distribution networks. Furthermore, studies are investigating the fundamental role of weight decay in training deep learning models used for line loss prediction, revealing its impact on optimization dynamics rather than solely as a regularization technique.

Papers

February 15, 2024

Criterion Collapse and Loss Distribution Control
Matthew J. Holland
Optimization Purpose Partial Monotonicity Gradient Descent Ascent Front Door Criterion Error Minimizing Noise Line Loss

December 19, 2023

Short-Term Multi-Horizon Line Loss Rate Forecasting of a Distribution Network Using Attention-GCN-LSTM
Jie Liu, Yijia Cao, Yong Li, Yixiu Guo, Wei Deng
Graph Convolutional Network Distribution System Multi Horizon Line Loss

October 6, 2023

Why Do We Need Weight Decay in Modern Deep Learning?
Francesco D'Angelo, Maksym Andriushchenko, Aditya Varre, Nicolas Flammarion
Implicit Regularization Regularization Effect Weight Decay Modern Deep Learning Line Loss

Line Loss

Papers

Criterion Collapse and Loss Distribution Control

Short-Term Multi-Horizon Line Loss Rate Forecasting of a Distribution Network Using Attention-GCN-LSTM

Why Do We Need Weight Decay in Modern Deep Learning?