强跟踪滤波器在管道泄漏检测与定位中的应用

摘要

结合管道流体瞬变特性,探讨了强跟踪滤波器在管道泄漏检测与定位中的应用.基于特征线法,流体瞬变过程的偏微分方程可以转化为差分方程形式.结合边界条件建立管道状态空间模型,以管线虚拟分段节点处的压力、流量、泄漏尺寸为增广的状态向量,建立管道的强跟踪滤波器(STF).通过引入时变渐消因子,实时调整状态预报误差协方差矩阵的大小,减弱管道的历史数据对滤波值的影响,从而提高滤波器对泄漏突变的强跟踪能力.实验结果表明STF算法可以实现快速和准确的管道泄漏检测与定位,将强跟踪滤波器应用于管道泄漏检测和定位是可行的.%Combining with pipeline fluid transient characteristics, the application of pipeline leak detection and location method is discussed by using strong tracking filter(STF). Based on characteristics method, the partial differential equations model of the pipeline is transformed into a set of difference equations. The state space model is constructed by combining with the boundary conditions. The pressure, flow rate and leakage size act as augmented state variables. The STF of the pipeline model is established. By introducing a time variance fading factor, the element values of forecast error covariance matrix is adjusted in time. The influence of historical data to established values is weakened and the leak tracking performance of the filter is improved. The results indicate that the STF can realize fast and accurate pipeline leak detection and location. It is feasible to apply STF on the pipeline leaking detection and location.