Researchon On-linemonitoring And Fault Diagnosis Of The Under Groundpipe Network

With the rapid development of our country’s economy and urbanization，variouscrisscrossed underground pipes constitute the complex pipe network system. Theunderground pipe network system plays a so important role thatmore and more peoplepay their attention to it.It can be divided into fluid pipeline network systemsuch as watersupply pipes, drainage pipes, gas pipes and so on, and no-fluid pipeline system such aspower cable. So far, the methods of underground pipeline network monitoring anddiagnoses are relatively lessandmost of them are focused on the off-line diagnoses onsingle long pipes.In response to research for underground pipeline network monitoring anddiagnoses, we select power cable and fluid type of water supply pipe as targets in thispaper.For on-line monitoring and fault diagnosis problem of the power cables, firstly, thispaperstudies the cable insulation fault of qualitative and discussesthe law of thezerosequence current signals when underground cable insulation fault occurred. Then acable simulation model is established and the simulation results show that this methodcan be sensitive to specific the grounding fault condition of cable. Secondly, this paperstudies the advantages of wavelet in cable fault traveling wave signal processing. We setup a single-phase ground fault、a two-phase ground fault and a single-phase break faultcable fault models in30km、50km and80km of the cable, and the simulation calculationresults show the method of the wavelet analysis can be accurate to determine the faultlocation of cable with the error less than5%.For on-line monitoring and fault diagnosis problem of water supply pipe networks,we study the stress testing point layout optimizationusing EPANET and fuzzy clusteringanalysis method, put forward a more optimized and reasonable layout method. In thisnew method the pressure testing point location is more accurate and greatly reducesuncertainty and redundancy. For the leakage of water supply pipe network fault location,this paper researches the theory of neural networks and applies it todeterminethe faultlocation. We set up a water supply pipe network leakage fault model in EPANET, and use the neural network to determine the fault location, results show that this methodbased on neural network for water supply pipe network leakage failure hasa very goodpositioning effect and its error is blow to2%.