论文部分内容阅读
通过对压力梯度法的分析,发现管道泄漏点的定位精度取决于摩阻系数。而传统的摩阻系数确定方法并不适用于实际运行的泄漏管道。鉴于此,该文提出基于BP神经网络预测摩阻系数的泄漏点定位方法。该方法以泄漏点前后的平均流量分别作为BP网路的输入单元预测泄漏点前后的摩阻系数,之后利用压力梯度法求解泄漏点位置。通过管道泄漏的水力模型实验验证了BP神经网络预测摩阻系数的有效性以及该方法应用于管道泄漏点定位的合理性。
Through the analysis of the pressure gradient method, it is found that the positioning accuracy of pipeline leakage depends on the friction coefficient. The traditional friction coefficient determination method is not suitable for the actual operation of the leaky pipe. In view of this, this paper presents a BP neural network prediction friction coefficient of the leakage point positioning method. The method uses the average flow before and after the leakage point as the input unit of the BP network respectively to predict the friction coefficient before and after the leakage point, and then uses the pressure gradient method to solve the leakage point position. The hydraulic model experiment of pipeline leakage verified the effectiveness of BP neural network in predicting the friction coefficient and the rationality of the method applied to locate the pipeline leak point.