基于相空间重构小波神经网络径流式水电站发电量预测

发布时间：2018-04-16 20:58

本文选题：发电量 + 径流式水电站　；参考：《中国农村水利水电》2017年09期

【摘要】：预测径流式水电站发量。针对径流式水电站发电量变化的随机性及水力发电系统的复杂非线性,提出一种基于相空间重构小波神经网络的径流式水电站发电量预测模型,并借助具有全局搜索能力的改进粒子群优化算法对小波神经网络特征参数进行优化,提高模型的精度和泛化能力。结果表明:相空间重构小波神经网络模型的预测平均相对误差约为8.7%,相关系数达到0.81,误差分析指标优于传统神经网络预测模型,在多步预测情形下模型的收敛性和稳定性得到较为明显增强,能够准确进行发电量预测。
[Abstract]:Prediction of runoff generation of hydropower stations.In view of the randomness of power generation of runoff hydropower station and the complex nonlinearity of hydropower system, a forecasting model of generating quantity of runoff hydropower station based on phase space reconstruction wavelet neural network is proposed.The improved particle swarm optimization algorithm with global searching ability is used to optimize the characteristic parameters of wavelet neural network to improve the accuracy and generalization ability of the model.The results show that the average relative error of the phase space reconstruction wavelet neural network model is about 8.7 and the correlation coefficient is 0.81.The error analysis index is better than the traditional neural network prediction model.The convergence and stability of the model are obviously enhanced in the case of multi-step prediction, which can accurately predict the generation capacity.
【作者单位】：西北农林科技大学水利与建筑工程学院;
【基金】：国家科技支撑计划项目(2012BAD10B02) 中央高校基本科研业务费专项资金(22050205) 陕西省水利科技项目(SLKJ-2013-14)
【分类号】：TV752

【相似文献】