摘要
针对电力市场提前一天市场清算电价预测,考虑历史负荷和历史清算电价对未来时段电价的影响,分别采用BP神经网络和遗传算法优化BP网络的方法,建立了提前一天电力市场清算电价预测模型。并采用澳大利亚昆士兰州电力市场1998年的电价数据分别进行了训练和预测,对结果进行了比较分析,结果表明遗传算法优化BP网络模型具有良好的预测效果。
The MCP(Market Clearing Price)forecast for day - ahead energy market considers the impact of the historical loads and MCPs on it,which applies BP neural network and genetic - algorithm - optimized neural network respectively to set MCP forecast models for day - ahead energy market.The historical loads and prices of Queensland energy market in1998are used for training and forecast testing.Preliminary results show that the neural network optimized by genetic algorithm is promising.
出处
《电力自动化设备》
EI
CSCD
北大核心
2004年第3期53-56,共4页
Electric Power Automation Equipment
关键词
提前一天市场
市场清算电价
神经网络
遗传算法
day - ahead market
market clearing price
neural network
genetic algorithm
