论文部分内容阅读
以各时刻的观测降深与流量的比值作为输入向量，含水层的导水系数T和储水系数S作为输出向量构成一个求算合水层参数的BP网络，用泰斯公式生成200个训练样本来训练网络，然后把一个实际的抽水试验资料加入到已训练好的网络中，输出的含水层参数（T和S）与配线法结果一致，展现了这种方法的实用性。
Taking the ratio of observed depth and flow at each moment as input vector, the aqueduct’s hydraulic conductivity T and water storage coefficient S are taken as output vector to form a BP network for calculating the parameters of the aquifer, and the TWIS formula is used to generate 200 trainings Samples were used to train the network, and then an actual pumping test data was added to the trained network. The output aquifer parameters (T and S) were consistent with those of the wiring method, demonstrating the practicability of this method.