Load forecasting using artificial neural networks

摘要

Artificial neural networks, modeled after their biological counterpart, have been successfully applied in many diverse areas including speech and pattern recognition, remote sensing, electrical power engineering, robotics and stock market forecasting. The most commonly used neural networks are those that gain knowledge from experience. Experience is presented to the network in the form of training data. Once trained, the neural network can recognize data that it has not seen before. This paper presents a fundamental introduction to the manner in which neural networks work and how to use them in load forecasting.< <ETX xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">&gt;</ETX>