基于无线传感器网络的自来水管道漏水监测定位关键技术研究

The leakage of water supply pipeline has become one of the most important cause of the waste of urban water resources and the pollution of the city.Therefore,it has arised more and more attention. The location of the leakage is the primary work for sovling the problem. With the rapid development and wide application of wireless sensor networks, it is possible to monitor and locate the leakage of liquid pipeline network by using network resources and effective communication technology. This project intends to study the key techniques of monitoring and locating water supply pipeline leakage based on wireless sensor networks. According to the distribution information of water supply pipeline and the abruptness features of the leakage, we intend to use improved TEEN algorithm to select hierarchical management structure and cluster-head for wireless sensor networks. Then the frequency feature of the leak signal is obtained by applying spectral subtraction. Furthermore, the time delay between different sensors which is used to locate the leakage is estimated by adaptive noise cancellation and adaptive time delay estimation algorithms. Finally, by making full use of the node resources of wireless sensor network and merging the multiple time delay with fusion algorithm, the accurate location of the leakage is achieved. The research results can not only be used for monitoring and locating the water supply pipeline leakage, but also providing valuable theoretical and technical support for monitoring and locating other leakage of liquid (such as natural gas, oil) in the underground pipe network.

自来水管道泄漏是城市水资源浪费的一个重要原因，同时给城市环保也带来了不良影响，因此治理管道漏损已引起高度重视，但要治理漏损，首先要确定漏水点。随着无线传感器网络的快速发展和广泛应用，利用网络资源和高效的通信手段进行流体管网泄漏的监测与定位成为了可能。本课题拟研究基于无线传感器网络的自来水管道漏水监测与定位的关键技术。根据供水管道分布及漏水事件的突发性特征，拟采用改进的TEEN算法进行网络分层管理及簇首选择；运用谱减法提取漏水信号特征信息，采用自适应噪声抵消和时延估计方法得到漏水点定位所需的多节点时延估计值；利用无线传感器网络的节点资源，选取合适的融合算法，通过对多节点时延估计值的融合处理，实现对漏水点的精确定位。本课题的研究成果不仅可用于自来水管道泄漏的监测与定位，还能为其他地下管网流体（如天然气、石油）泄漏的监测与定位提供有价值的理论和技术支持。

供水管道的泄漏会造成水资源的巨大损失，研究供水管道泄漏检测定位方法对于保证供水管网安全运行和节约水资源具有重要意义。针对传统管道泄漏检测定位方法智能化程度和检测效率较低的问题，本项目组对基于无线传感器网络的管道泄漏检测与定位系统和方法开展了研究。首先开发了基于ZigBee的供水管道远程监测系统。该系统通过布置在管道上的无线传感器进行数据采集，通过GPRS和WiFi网络实现数据远程传输。针对传统无线传感器网络协议功耗高的问题，提出了一种漏水触发组网方法和高能效LEACH路由协议，能够有效降低无线传感器网络的能耗，提高网络的生存周期。仿真分析与实验结果表明，我们提出的泄漏识别方法能够有效的识别出管道的泄漏情况。本文提出的漏水触发组网方法与传统无线传感器网络组网方法相比具有更低的能量消耗。为了提高管道泄漏检测的精度和智能化程度，我们进一步研究了基于信号时频特征和BP神经网络的泄漏识别方法。该方法利用谱减法进行信号增强，然后将多个时频特征输入BP神经网络来判断是否发生泄漏。在此基础上我们进一步提出了基于信号时频特征和支持向量机的泄漏识别方法。该方法利用固有模态函数、近似熵、主成分分析构建信号特征集，采用支持向量机作为分类器实现泄漏检测。对于发生泄漏的情况进一步开发了基于分数低阶自适应时延估计的漏点定位方法，从而实现了管道状态监测和泄漏识别与漏点定位。