State Analysis And Anomaly Detection Application Of Water Supply Pipeline Network System

Water Supply Network system is an important part of municipal infrastructure,and it has a direct impact on people's industrialproduction and dairylife.Getting status of the pipeline network accuratly is very important to the daily management and anomaly detection of water supply enterprises.However,there are so many uncertainties affect the result of state analysis in water supply network system.Therefore,based on the analysis of the existing research,this paper carries out the following research about the acquisition of status and anomaly detection of pipeline network system:(1)Hydraulic analysis of water supply network system considering uncertainty.Considering of the uncertainties of demanding and pipeline friction,this paper analyzes the influence of uncertainties on the fluctuation of node pressure in the water supply network system by Monte Carlo simulation method.Subsequently,this paper analyzes the influence of operating conditions,the number of monitoring points and the spatial location on the node pressure quantification as well.The result shows that the fluctuation of node pressure is positively correlated with the degree of uncertainty,and arranging monitors reasonablely is good for evaluating accurately.(2)Modeling and analyzing of water supply network based on state space.Considering that the urban water supply network system has the characteristics of complex structure and huge scale,a state space modeling method based on VARX model is proposed in this paper.Combined with the operation data of the example pipeline network,the state space expression of the system is obtained by VARX modeling,differential and Z-transformation.This method has been proved to be feasible by comparing the system matrix established by this method with the state space modeling method based on the system mechanism model of water supply network.(3)Considering that the synthetic data obtained from EPANET hydraulic model or the estimated data obtained from the Soft-sensing will be affected by uncertainty,and the SPC method is a statistical analysis method based on operational state data,this chapter attempts to discuss how to use the SPC method under these circumstances.If the interested nodes are equipped with pressure monitoring equipments,the corresponding SPC control chart can be drawn directly by the measured data from monitor.If the interested nodes are not equipped with pressure monitoring equipments,it is necessary to use Monte Carlo simulation method to calculate the average and standard deviation of simulation values at each time,taking into account the constraints of monitoring points.And,in order to reduce the false-alarm rates of detecting,this paper uses SPC method to detect anomalies by pressure data from the minimum water use period at night and the stable water use period.(4)State-space modeling and observability application of water supply network.According to the research results of state space modeling of water supply network system,the state space modeling method of water supply network based on VARX model is used to realize the state space modeling of actual water supply network system,and the Observability analysis,the Minimum realization and the influence of system order on observability are studied.Then,on the basis of studying the influence of renovation,expansion and explosion events on the state space model,the application of the state space model and its observability in anomalous event detection of water supply network system is discussed.And on the basis of studying the influence of renovation and expansion and explosion on state space model,the application of state space model and observability of water supply network system in anomaly detection is discussed in the paper.