自适应神经网络在短期负荷预测中的应用

采用基于混沌算法的自适应预测模型,应用于电力系统短期负荷预测.选取重构相空间中的饱和嵌入维数作为神经网络的输入节点数,适当选择非线性反馈项,能使网络的动力学在权空间具有混沌行为.通过进化算法建立一种自适应机制,使得网络能够根据学习和训练的结果优化非线性反馈项.算例表明,该算法具有很强的自适应能力和鲁棒性,精度高.

The Application of Adaptive Neural Network in Short Term Load Forecasting

An adaptive prediction model based on chaotic algorithm was applied to short term load forecast of power system. Taking the saturation inset dimension of reconstructed phase space as the input node number of artificial neural network and suitable nonlinear feedback terms are selected, the dynamics of network become chaotic in the weight space. EP evolutionary computation is used to establish a kind of self-adaptive prediction model by which the nonlinear feedback term is optimized according to the outcome of learning and training. It is revealed that the algorithm has good performance in self-adaptive ability, robustness and precision.