粒子群小波人工神经网络组合模型的径流预测

Monthly Runoff Forecasting by WANN Model based on PSO

为了克服传统径流过程预测容易产生累积误差的缺点,提高径流预测精度,提出了一种基于粒子群小波人工神经网络组合模型的月径流过程预测算法,该算法具有原理简单、实用性强等特点。将该算法用于预测某电厂月径流过程计算,结果表明,其预测结果精度高,可为水电厂提供可靠的入库径流,对水电厂制定合理的运行方式有重要作用。

In order to overcome the shortcoming of easily-produced accumulation error of tradition monthly runoff process forecasting and to improve the accuracy of runoff forecasting, this paper puts forward a monthly runoff forecasting algorithm based on a WANN model based on PSO. This algorithm is easy in principle and good in practicality. This algorithm was applied to the monthly runoff forecasting of a hydropower plant and the results showed that its forecasting accuracy is high, so that it can provide reliable inflow for hydropower station and plays a great role in making reasonable operation modes of hydropower statio.