基于适应性子集度神经网络的电力系统短期负荷预测  

Short-term load forecasting method based on adaptive subsethood theory of neural network

作　　者：刘耀年[1] 王振浩[1] 曾令全[1] 李月玲[1] 

机构地区：[1]东北电力学院,吉林吉林132010

出　　处：《电工电能新技术》2003年第4期6-9,52,共5页Advanced Technology of Electrical Engineering and Energy

摘　　要：依据模糊模式识别、子集度测度理论,提出一种适应性子集度测度神经网络的短期负荷预测的新方法;该方法可在线地学习神经网络的结构和参数,确定连接方式和神经元节点。仿真结果表明:该方法具有预测精度高、速度快的优点,是值得广泛推广的好方法。A short\|term load forecasting technique for power system based on adaptive subsethood theory of neural network is presented, which uses fuzzy pattern recognition and adaptive subsethood theory. The algorithm can adjust network configuration and parameters on line, defines joints and nodes of neural network. Theoretical analysis and computer simulation results demonstrate that the forecasting method has the advantages of higher forecasting accuracy and smaller forecasting error, and the method may be popularized. 

关 键 词：神经网络 子集度测度理论 短期负荷预测 

分 类 号：TM734[电气工程—电力系统及自动化]