【摘要】 为了解决管道泄漏检测与定位问题,采用基于粒子群算法的管道泄漏模型反问题方法求解泄漏点大小和位置,得到该方法在管道参数波动情况下的鲁棒性结果,对管道参数波动进行敏感度分析.由于管道泄漏模型采用偏微分方程描述,给出该模型求解的初始条件和边界条件,根据这些条件和已有实验平台仿真泄漏模型的稳态和动态状况.基于管道泄漏模型,对达西-威斯巴哈摩擦系数f和泄漏小孔的流通系数C1进行敏感度分析.在人为地增加参数扰动后,采用粒子群算法进行反问题求解.从搜索结果可以看出,参数的敏感性越强,粒子群算法对参数的鲁棒性越弱.更多还原

【Abstract】 The detection of the pipeline leakage can be formulated as an inverse problem which is solved by using the particle swarm optimization(PSO)method.The sensitivity analysis was conducted with respect to various pipeline parameters(such as Darcy-Weisbach friction factor fand leakage holes flow coefficient C1,etc.).The results of robust corresponding to these parameters were given.A classical pipeline model governed by nonlinear hyperbolic partial differential equations(PDEs)was employed to obtain a numerical simulation based on the experimental platform.Then the PSO algorithm was used to search leakage parameters and perturbations were introduced to implement sensitivity analysis.Results demonstrate that the particle swarm algorithm’s robustness towards parameters decreases with the parameter’s sensitivity increases.更多还原