摘要
针对ＢＰ网络的缺陷，提出了基于拟牛顿法的自适应算法和改进的遗传算法，以提高神经网的学习效率，克服ＢＰ网络的局部收敛性的缺点，形成一种新的神经网与遗传算法相结合的短期负荷预测模型。
This paper presents a new hybrid artificial neural network (ANN) learning algorithm, which combines the adapted learning algorithm based on quasi Newton method and the improved genetic algorithms (GA) It possesses merits of high speed of convergence and higher precision A new hybrid short term load forecasting model is built with the above algorithms Satisfactory results are obtained by applying it to a certain network
出处
《电力系统自动化》
EI
CSCD
北大核心
1997年第3期29-32,共4页
Automation of Electric Power Systems
关键词
人工神经网
遗传算法
负荷预测
电力系统
artificial neural network (ANN) genetic algorithm (GA) short term load forecasting the adapted learning algorithm based on quasi Newton method
