The Research Of Localization Of Fault Of The Water Supply Network

This topic's core research content was the localization of the pipe explosion of the urban water supply network. During the research process, we have done the simulating experiment of the pipe explosion of the water supply network. At the same time, we established the neutral network model of the detection and localization of the pipe explosion in order to achieve the goal of the quick localization of fault.The first, the author did some simple introduction of the nowadays localization of the pipe explosion according to the national method and influencing factor of the localization of the pipe explosion. The author also pointed out the existing problem and put forward the thought of the localization of the pipe explosion based on neutral network.The second, during the simulation of the localization of fault in the lab, we have done the layout of the supervision point based on the pressure distribution. We calculate water peak matrix, then we standardize the matrix according to the standardization method and got the fuzzy similar matrix. At last, we used minnow net method to obtain the fuzzy equivalence matrix. So we have classified all the nodal points.The third, we used neutral net to deal with the data that we obtained from the experiment. Also, we predicted the virtual pipe explosion. And, the implication relation of"pipe explosion of any point"and real-time operation mode had been reflected. Then, we add some buzz into data which was got from the real-time operation mode. After that, we input the data into the neutral net we had trained before to verify the accurate of it.The last, we compiled the software of the localization of the pipe explosion. The functions of neutral net calculation, manual prediction, modification of parameters, system configuration and etc. had been accomplished. Also, the function of switching on-line supervision and manual supervision is also been designed. For the extensive use of the software, we also installed the function of the prediction of the other neutral net according to other neutral net. We had one group of data to test the module, the result was satisfaction.