输水线路旁侧出流预测的数值研究

Numerical simulation of water loss prediction along the water diversion way

调水、引水工程沿线非计划用水的评估计算及控制是受水区能否得到足够水量的关键问题。选取南水北调工程苏北江都至淮安调水线路进行数值试验,采用神经网络技术预测非计划旁侧出流,探讨该技术的精度和可靠性。根据数值实验的结果,和传统水力学方法的计算结果相比较,神经网络模型的预测、预报精度较高,但用于实际预测还需实测数据作进一步验证。

The evaluation and control of unforeseen water consumption along the way of water diversion projects is one of the key problems to ensure that enough water can be transferred to the destination. Artificial Neural Network (ANN) is applied to predict unforeseen water consumption along the way from Jiangdu to Huai'an as a case, which is a part of the South-to-North Water Transfer Project, and the precision and reliability of the method are discussed. Results of numerical experiment show that the accuracy of ANN prediction is higher than that of conventional hydraulic methods, while more field data is needed for verification when ANN is used in practical projects.