论文部分内容阅读
将神经网络非线性学习能力与小波变换多尺度分解特性相结合,建立太阳逐日辐射能量小波神经网络预测模型。通过小波多尺度分解使太阳逐日辐射能量序列在一定尺度上表现出准平稳性,以太阳逐日辐射能量相空间重构数据作为模型的输入量对网络进行训练。仿真结果表明该方法可较好地用于太阳逐日辐射能量预测。
The neural network nonlinear learning ability and wavelet transform multi-scale decomposition characteristics are combined to establish solar day-to-day radiant energy wavelet neural network prediction model. The solar day-to-day radiative energy sequence is quasi-stationary at a certain scale by wavelet multi-scale decomposition. The network is trained by the solar day-to-day radiative energy phase space reconstruction data as the input of the model. The simulation results show that this method can be used to predict solar daily solar radiation.