CN112097125A - Water supply pipe network pipe burst detection and positioning method based on self-adaptive checking - Google Patents

Water supply pipe network pipe burst detection and positioning method based on self-adaptive checking Download PDF

Info

Publication number
CN112097125A
CN112097125A CN202010987621.9A CN202010987621A CN112097125A CN 112097125 A CN112097125 A CN 112097125A CN 202010987621 A CN202010987621 A CN 202010987621A CN 112097125 A CN112097125 A CN 112097125A
Authority
CN
China
Prior art keywords
pipe
moment
node
detected
water demand
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Granted
Application number
CN202010987621.9A
Other languages
Chinese (zh)
Other versions
CN112097125B (en
Inventor
周啸
信昆仑
徐玮榕
陶涛
李树平
颜合想
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Tongji University
Original Assignee
Tongji University
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Tongji University filed Critical Tongji University
Priority to CN202010987621.9A priority Critical patent/CN112097125B/en
Publication of CN112097125A publication Critical patent/CN112097125A/en
Application granted granted Critical
Publication of CN112097125B publication Critical patent/CN112097125B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • F—MECHANICAL ENGINEERING; LIGHTING; HEATING; WEAPONS; BLASTING
    • F17—STORING OR DISTRIBUTING GASES OR LIQUIDS
    • F17D—PIPE-LINE SYSTEMS; PIPE-LINES
    • F17D5/00—Protection or supervision of installations
    • F17D5/02—Preventing, monitoring, or locating loss
    • F17D5/06—Preventing, monitoring, or locating loss using electric or acoustic means
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/24—Classification techniques
    • G06F18/243—Classification techniques relating to the number of classes
    • G06F18/2433—Single-class perspective, e.g. one-against-all classification; Novelty detection; Outlier detection
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
    • G06Q50/06—Energy or water supply

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Business, Economics & Management (AREA)
  • Theoretical Computer Science (AREA)
  • Data Mining & Analysis (AREA)
  • Economics (AREA)
  • Health & Medical Sciences (AREA)
  • General Physics & Mathematics (AREA)
  • General Engineering & Computer Science (AREA)
  • Evolutionary Biology (AREA)
  • Marketing (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Bioinformatics & Computational Biology (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • Artificial Intelligence (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Public Health (AREA)
  • Water Supply & Treatment (AREA)
  • General Health & Medical Sciences (AREA)
  • Human Resources & Organizations (AREA)
  • Evolutionary Computation (AREA)
  • Primary Health Care (AREA)
  • Strategic Management (AREA)
  • Tourism & Hospitality (AREA)
  • General Business, Economics & Management (AREA)
  • Acoustics & Sound (AREA)
  • Mechanical Engineering (AREA)
  • Pipeline Systems (AREA)
  • Examining Or Testing Airtightness (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)

Abstract

The invention relates to a water supply network pipe burst detecting and positioning method based on self-adaptive checking, which comprises the following steps: (1) collecting pressure and flow monitoring data of a water supply network, and estimating water demand of each node by using a self-adaptive check algorithm; (2) processing the water demand checking result to construct a detection cluster; (3) judging whether each detection cluster is an outlier or not by adopting a DBSCAN algorithm; (4) and analyzing the outlier based on the pipe network flow response characteristics during pipe explosion, outputting a pipe explosion detection result, and judging the position of the pipe explosion. Compared with the prior art, the invention has the advantages of high detection and positioning precision, small influence by model and data uncertainty and low cost.

Description

Water supply pipe network pipe burst detection and positioning method based on self-adaptive checking
Technical Field
The invention relates to a method for detecting and positioning burst pipes of a water supply network, in particular to a method for detecting and positioning burst pipes of a water supply network based on self-adaptive checking.
Background
Leakage control is a key problem in the operation and maintenance of water supply network systems. Leakage not only can cause unnecessary waste of water resources and increase the water supply cost, but also can cause the consumption of additional energy and chemical agents in the water treatment process and increase the risk that a pipe network is polluted by microorganisms and pollutants. Pipe explosion is one of important components and expressions of pipe network leakage; in addition, pipe bursting can cause problems such as water supply interruption, ground settlement, etc. Due to the fact that factors causing pipe explosion are very various, the pipe explosion is difficult to completely avoid in practical engineering. Therefore, the pipe explosion can be timely detected and positioned during pipe explosion, and the pipeline can be quickly repaired, so that the effective way for improving the water supply service quality, reducing the leakage rate and reducing the water supply cost is realized. Partial pipe explosion can cause water flow to overflow the ground and even the residents cut off the water, so that the water can be detected through manual observation or complaints of the residents around, and the water can be repaired in time by a water supply department. Nevertheless, the pipe explosion detection efficiency can be obviously improved through the automatic pipe explosion detection and positioning means, and especially for the pipe explosion which occurs at night or in remote areas in cities and towns, the time for discovering the pipe explosion can be greatly shortened. Because large pipe explosion can cause serious influences such as large water resource loss, damage of peripheral facilities and the like in a short time, it is very important to improve the pipe explosion repairing efficiency by adopting a proper pipe explosion detecting and positioning algorithm.
A great deal of research on a water supply network pipe burst detection method is already carried out at home and abroad. The existing pipe burst detection and positioning method can be mainly divided into a data driving method, a hydraulic model method, an acoustic analysis method, a transient flow analysis method and the like. The following are some representative studies:
1) data driving method
As in the literature:
[1]Wu Y.,Liu S.,Wu X.,Liu Y.,Guan Y.Burst detection in district metering areas using a data driven clustering algorithm.Water Research,2016,Vol.100:28-37.
[2]Mounce S.R.,Khan A.,Wood A.S.,Day A.J.,Widdop P.D.,Machell J.Sensor-fusion of hydraulic data for burst detection and location in a treated water distribution system.Information Fusion,2003,Vol.4(3):217-229.
the method adopts the following main technical measures: and (3) detecting and positioning the pipe explosion according to the difference between the monitoring data during pipe explosion and the monitoring data under the normal operation working condition by analyzing the monitoring data time and/or spatial distribution characteristics of a supervisory control and data acquisition (SCADA) system of the pipe network.
The advantages and disadvantages are as follows: the method has the advantages that the information in the monitoring data can be directly analyzed, the method does not depend on and use an accurate pipe network hydraulic model, the theoretical basis is concise, and the method is easy to use. However, such methods have the following disadvantages: (1) the method is easily influenced by factors such as monitoring errors and normal water consumption fluctuation in monitoring data, and the situations of low detection rate, high false alarm rate and the like occur; (2) the hydraulic characteristics of the pipe network are ignored, so that the positioning accuracy is limited; (3) the use of a part of data driving methods depends on the construction of independent metering areas (DMA) of the pipe network, and is difficult to be directly applied to most of the pipe networks in China.
2) Hydraulic model method
As in the literature:
[3]Wu Z.Y.,Sage P.,Turtle D.Pressure-dependent leak detection model and its application to a district water system.Journal of Water Resources Planning and Management,2010,Vol.136(1):116-128.
[4]Meseguer J.,Mirats-Tur J.M.,Cembrano G.,Puig V.,Quevedo J.,Pérez R.,Sanz G.,Ibarra D.A.decision support system for on-line leakage localization.Environmental Modelling&Software,2014,Vol.60:331-345.
the method adopts the following main technical measures: and analyzing the monitoring data by using a pipe network hydraulic model, and solving unknown quantities such as the position, the size and the like of pipe burst (leakage) in the pipe network by an optimization method. The difference between the model simulation result after adding a pipe burst (leak) and the actual monitored value is generally taken as an objective function.
The advantages and disadvantages are as follows: the method has the advantages that pipe burst detection and positioning can be realized at the same time, and higher detection and positioning precision can be obtained when the information such as a pipe network hydraulic model, monitoring data and the like is accurate. But has the following disadvantages: (1) the method depends heavily on the accuracy of the pipe network model, and because a large amount of manpower and material resources are needed for accurately checking the pipe network model, the practical application scene of the method is limited; (2) the method is easily influenced by factors such as monitoring errors and normal water consumption fluctuation in monitoring data.
3) Acoustic analysis method
As in the literature:
[5]Kang J.,Park Y.,Lee J.,Wang S.,Eom D.Novel leakage detection by ensemble CNN-SVM and graph-based localization in water distribution systems.IEEE Transactions on Industrial Electronics,2018,Vol.65(5):4279-4289.
the method adopts the following main technical measures: the noise generated by the leakage point of the pipe network is detected and analyzed by technical workers through devices such as a leakage listening rod and a correlator so as to carry out detection and positioning. In recent years, there have been researches to automatically analyze the missing point noise by using an algorithm for detection and positioning.
The advantages and disadvantages are as follows: the method is convenient to use and high in positioning accuracy, and is a method widely adopted by the water supply department at present. However, this method requires a lot of manpower, and the acoustic signal is attenuated and disturbed after propagating a certain distance, so that the use cost is high, and it is difficult to determine the occurrence of the pipe burst directly from a large area.
4) Transient flow analysis method
As in the literature:
[6]Lee P.J.,Lambert M.F.,Simpson A.R.,Vítkovsky J.P.,Misiunas D.Leak location in single pipelines using transient reflections.Australasian Journal of Water Resources,2007,Vol.11(1):53-65.
[7]Covas D.,Ramos H.Case studies of leak detection and location in water pipe systems by inverse transient analysis.Journal of Water Resources Planning and Management,2010,Vol.136(2):248-257.
the method adopts the following main technical measures: the detection and the positioning of pipe network pipe explosion points are carried out by analyzing the propagation, reflection and attenuation of transient flow caused by pipe explosion in a pipe network or analyzing the change of the transient flow generated when the transient flow passes through a damaged position of a pipeline.
The advantages and disadvantages are as follows: the method has higher precision for positioning the pipe explosion point of the pipe network, but has the following defects: (1) the calculation is complex, most of the current researches can only be applied to a single pipeline or a simple pipe network, and the requirement on the calculation capacity is sharply increased when the pipe network of a target area is complex; (2) a high-frequency pressure signal collector needs to be installed, and the data transmission and storage cost is high; (3) the transient flow signals are easily interfered by factors such as unknown pipe sections and blockage in the pipe network. Therefore, the method is difficult to be directly applied to large and complex pipe networks in practical engineering.
In summary, the water supply mechanism in our country still relies heavily on detecting and positioning the pipe burst by manual inspection. Although a great deal of research on pipe burst detection and positioning is available, the pipe burst detection and positioning method is limited by the pipe network hydraulic model precision, the monitoring data error, the cost and other factors, and the pipe burst detection and positioning method is still lack at present, and is high in detection and positioning precision, small in influence of model and data uncertainty and low in cost.
Disclosure of Invention
The invention aims to overcome the defects in the prior art and provide a water supply network pipe burst detecting and positioning method based on adaptive check, which has high detecting and positioning accuracy, small influence by model and data uncertainty and low cost.
The purpose of the invention can be realized by the following technical scheme:
a water supply network pipe burst detection and positioning method based on self-adaptive check comprises the following steps:
(1) collecting pressure and flow monitoring data of a water supply network, and estimating water demand of each node by using a self-adaptive check algorithm;
(2) processing the water demand checking result to construct a detection cluster;
(3) judging whether each detection cluster is an outlier or not by adopting a DBSCAN algorithm;
(4) and analyzing the outlier based on the pipe network flow response characteristics during pipe explosion, outputting a pipe explosion detection result, and judging the position of the pipe explosion.
Preferably, step (1) is specifically:
(11) obtaining pipe network monitoring data toLess including n before the moment to be detecteddMonitoring pressure and flow at each moment in the day;
(12) grouping pipelines and nodes in a water supply network respectively, and setting the roughness coefficients of all the pipelines in the same group to be the same and the daily water consumption change curves of all the nodes in the same group to be the same;
(13) and inputting the pressure and flow monitoring data and the grouping information into a self-adaptive checking algorithm, and calculating the water demand value of the node at each moment in real time.
Preferably, step (13) is specifically:
(131) checking the node water demand at the current moment according to the rough coefficient at the previous moment, the pressure at the current moment and the flow monitoring data to obtain a reasoning observation value of the node water demand at the current moment;
(132) checking the pipeline rough coefficient at the current moment according to the inference observation value of the node water demand at the current moment, the pressure at the current moment and the flow monitoring data to obtain the inference observation value of the pipeline rough coefficient at the current moment;
(133) and calculating the optimal estimated values of the rough coefficient and the node water demand according to the rough coefficient and the node water demand at the previous moment, the rough coefficient and the node water demand at the current moment and the inference observation value of the node water demand by using a Kalman filter algorithm, and taking the optimal estimated values as final checking results of the rough coefficient and the node water demand at the current moment.
Preferably, the step (2) is specifically:
(21) defining adjacent nodes with upper and lower relations to be grouped into a detection cluster, wherein each detection cluster comprises p node groups, and further dividing the water supply network into a plurality of detection clusters;
(22) suppose the time to be detected is the nthdAt time t of day, any one detection cluster EkThe following matrixes to be detected are respectively constructed:
Figure BDA0002689773920000051
Figure BDA0002689773920000052
to detect clusters EkThe ith node is grouped into a node water demand checking column vector at the time t,
Figure BDA0002689773920000053
element (1) of
Figure BDA0002689773920000054
Indicating a detection cluster EkThe ith node is grouped into the node water demand checking list result at the time t of the jth day, i is 1, 2 … …, p, j is 1, 2 … …, nd；
(23) For Qt_kEach column of
Figure BDA0002689773920000055
Order to
Figure BDA0002689773920000056
Wherein, muQIs a vector
Figure BDA0002689773920000057
Mean value of all elements in, σQIs a vector
Figure BDA0002689773920000058
Standard deviation of all elements in (a);
(24) rewriting the matrix to be detected as
Figure BDA0002689773920000059
Preferably, step (3) is specifically:
(31) determining algorithm parameters and gamma, wherein the parameters are the neighborhood radius of the DBSCAN algorithm, and gamma is the minimum clustering example number of the DBSCAN algorithm;
(32) judging whether the moment to be detected is an outlier or not by using a DBSCAN algorithm;
(33) and (4) if the moment to be detected is the outlier, otherwise, judging that the tube explosion does not occur at the moment to be detected.
Preferably, γ in step (31) is determined as: γ is 2 × p.
Preferably, step (31) is determined by:
to-be-detected matrix NQt_kEach line of (a) is respectively regarded as one point in the space, the Euclidean distance between every two points is respectively calculated, and n is obtained in totald(nd-1)/2 pairs of Euclidean distances, and arranging all the solved Euclidean distances from large to small to obtain a vector
Figure BDA00026897739200000510
Selecting a distance threshold xi, and ordering:
Figure BDA00026897739200000511
wherein round is an operation function, which means rounding to find the whole,
Figure BDA00026897739200000512
representing the vector dpTo middle
Figure BDA0002689773920000061
Data of individual elements.
Preferably, step (32) is specifically:
(321) to-be-detected matrix NQt_kEach line of the cluster mark C is respectively regarded as a point in the space, one point is randomly selected as a current point, and the cluster mark C is made to be 1;
(322) all the points with Euclidean distance smaller than the current point are taken as the adjacent points of the current point, if the number of the adjacent points is less than gamma, the current point is marked as an outlier, and the step (324) is carried out; otherwise, marking the current point as a core point of the cluster C;
(323) repeating the step (322) for each adjacent point of the current point in sequence until no new adjacent point appears and all the discovered adjacent points belong to the cluster C;
(324) and selecting the next unmarked point as the current point, marking the clustering mark C as C +1, repeating the steps (322) and (323), and repeatedly executing the steps until all the points are marked.
Preferably, step (4) is specifically:
(41) comparing the water demand checking result of each group in the outlier with the value at the same moment in the historical data, if the water demand of each group is smaller than the average value of the water demand at the same moment in the historical data, the qualitative analysis of the pipe bursting flow fails, and judging that the pipe bursting does not occur at the moment to be detected; otherwise, the pipe bursting flow qualitative analysis is passed, the pipe bursting at the time to be detected is judged, and the step (42) is executed;
(42) and regarding all detection clusters which are identified as outliers at the moment to be detected and pass through the qualitative analysis of the pipe burst flow, and regarding the area where the node group with the largest water demand rise compared with historical data in the detection clusters is located as the pipe burst occurrence position.
Preferably, after the step (41) of qualitatively analyzing the tube bursting flow, before the step (42) is executed, the method further comprises the step (41 a): if the outlier is not prompted again within T hours after the moment to be detected, the tube explosion continuity analysis is failed, and it is judged that tube explosion does not occur at the moment to be detected; if the outlier is prompted for the second time within T hours and the outlier passes the tube explosion flow qualitative analysis in the step (41), the tube explosion continuity analysis is judged to pass, the tube explosion occurs, the step (42) is entered,
further, the step (42) is specifically: and (4) regarding all detection clusters of which the moment to be detected is identified as an outlier and which pass through the pipe bursting flow qualitative analysis in the step (41) and the pipe bursting continuity analysis in the step (41a), regarding the area where the node group with the water demand rising most compared with the historical data in the detection clusters is located as the pipe bursting occurrence position.
Compared with the prior art, the invention has the following advantages:
(1) the method carries out pre-analysis on the monitoring data through self-adaptive check, carries out outlier detection on the unknown node water demand after estimating the unknown node water demand through the monitoring value, does not depend on using an accurate pipe network model to analyze the monitoring value, does not need to set accurate pipe network model parameters in advance, and has good engineering practicability.
(2) The invention greatly reduces the influence caused by the monitoring error of the pipe network by means of the self-adaptive checking technology, so that the pipe burst detection and positioning are more accurate.
(3) The invention can detect the pipe burst and determine the node grouping area where the pipe burst is positioned; in addition, through the analysis of the continuity of the tube explosion, the false alarm rate can be greatly reduced, and higher detection precision can be obtained under the same false alarm rate.
Drawings
FIG. 1 is a flow chart of a method for detecting and positioning water supply network pipe burst based on adaptive check according to the present invention;
FIG. 2 is a diagram of the pipe network structure, grouping, pipe bursting position and pipe bursting daily part pressure and flow monitoring value of an exemplary water supply pipe network in the embodiment;
fig. 3 (a) shows the result of checking the water demand for 24 hours on the day of the occurrence of pipe explosion, and fig. 3 (b) shows the result of checking the water demand for 15 days in the past at the same time of the pipe explosion.
Detailed Description
The invention is described in detail below with reference to the figures and specific embodiments. Note that the following description of the embodiments is merely a substantial example, and the present invention is not intended to be limited to the application or the use thereof, and is not limited to the following embodiments.
Examples
As shown in fig. 1, a method for detecting and positioning burst pipes of a water supply network based on adaptive check comprises the following steps:
(1) collecting pressure and flow monitoring data of a water supply network, and estimating water demand of each node by using a self-adaptive check algorithm;
(2) processing the water demand checking result to construct a detection cluster;
(3) judging whether each detection cluster is an outlier or not by adopting a DBSCAN algorithm;
(4) and analyzing the outlier based on the pipe network flow response characteristics during pipe explosion, outputting a pipe explosion detection result, and judging the position of the pipe explosion.
The step (1) is specifically as follows:
(11) acquiring pipe network monitoring data, at least including n before the moment to be detecteddMonitoring pressure and flow at each moment in the day;
(12) grouping pipelines and nodes in a water supply network respectively, and setting the roughness coefficients of all the pipelines in the same group to be the same and the daily water consumption change curves of all the nodes in the same group to be the same;
(13) and inputting the pressure and flow monitoring data and the grouping information into a self-adaptive checking algorithm, and calculating the water demand value of the node at each moment in real time.
The step (13) is specifically as follows:
(131) checking the node water demand at the current moment according to the rough coefficient at the previous moment, the pressure at the current moment and the flow monitoring data to obtain a reasoning observation value of the node water demand at the current moment;
(132) checking the pipeline rough coefficient at the current moment according to the inference observation value of the node water demand at the current moment, the pressure at the current moment and the flow monitoring data to obtain the inference observation value of the pipeline rough coefficient at the current moment;
(133) and calculating the optimal estimated values of the rough coefficient and the node water demand according to the rough coefficient and the node water demand at the previous moment, the rough coefficient and the node water demand at the current moment and the inference observation value of the node water demand by using a Kalman filter algorithm, and taking the optimal estimated values as final checking results of the rough coefficient and the node water demand at the current moment.
The step (2) is specifically as follows:
(21) defining adjacent nodes with upper and lower relations to be grouped into a detection cluster, wherein each detection cluster comprises p node groups, p is more than or equal to 3 and less than or equal to 5 generally, and one node group can be contained in a plurality of detection clusters, so that the water supply network is divided into a plurality of detection clusters;
(22) suppose the time to be detected is the nthdAt time t of day, any one detection cluster EkThe following matrixes to be detected are respectively constructed:
Figure BDA0002689773920000081
Figure BDA0002689773920000082
to detect clusters EkThe ith node is grouped into a node water demand checking column vector at the time t,
Figure BDA0002689773920000083
element (1) of
Figure BDA0002689773920000084
Indicating a detection cluster EkThe ith node is grouped into the node water demand checking list result at the time t of the jth day, i is 1, 2 … …, p, j is 1, 2 … …, nd；
(23) For Qt_kEach column of
Figure BDA0002689773920000085
Order to
Figure BDA0002689773920000086
Wherein, muQIs a vector
Figure BDA0002689773920000091
Mean value of all elements in, σQIs a vector
Figure BDA0002689773920000092
Standard deviation of all elements in (a);
(24) rewriting the matrix to be detected as
Figure BDA0002689773920000093
The step (3) is specifically as follows:
(31) determining algorithm parameters and gamma, wherein the parameters are the neighborhood radius of the DBSCAN algorithm, and gamma is the minimum clustering example number of the DBSCAN algorithm;
(32) judging whether the moment to be detected is an outlier or not by using a DBSCAN algorithm;
(33) and (4) if the moment to be detected is the outlier, otherwise, judging that the tube explosion does not occur at the moment to be detected.
In the step (31), gamma is determined as: γ is 2 × p.
The determination in step (31) is made as follows:
to-be-detected matrix NQt_kEach line of (a) is respectively regarded as one point in the space, the Euclidean distance between every two points is respectively calculated, and n is obtained in totald(nd-1)/2 pairs of Euclidean distances, and arranging all the solved Euclidean distances from large to small to obtain a vector
Figure BDA0002689773920000094
Selecting a distance threshold xi, and ordering:
Figure BDA0002689773920000095
wherein round is an operation function, which means rounding to find the whole,
Figure BDA0002689773920000096
representing the vector dpTo middle
Figure BDA0002689773920000097
Data of individual elements.
The step (32) is specifically as follows:
(321) to-be-detected matrix NQt_kEach line of the cluster mark C is respectively regarded as a point in the space, one point is randomly selected as a current point, and the cluster mark C is made to be 1;
(322) all the points with Euclidean distance smaller than the current point are taken as the adjacent points of the current point, if the number of the adjacent points is less than gamma, the current point is marked as an outlier, and the step (324) is carried out; otherwise, marking the current point as a core point of the cluster C;
(323) repeating the step (322) for each adjacent point of the current point in sequence until no new adjacent point appears and all the discovered adjacent points belong to the cluster C;
(324) and selecting the next unmarked point as the current point, marking the clustering mark C as C +1, repeating the steps (322) and (323), and repeatedly executing the steps until all the points are marked.
The step (4) is specifically as follows:
(41) comparing the water demand checking result of each group in the outlier with the value at the same moment in the historical data, if the water demand of each group is smaller than the average value of the water demand at the same moment in the historical data, the qualitative analysis of the pipe bursting flow fails, and judging that the pipe bursting does not occur at the moment to be detected; otherwise, the pipe bursting flow qualitative analysis is passed, the pipe bursting at the time to be detected is judged, and the step (42) is executed;
(42) and regarding all detection clusters which are identified as outliers at the moment to be detected and pass through the qualitative analysis of the pipe burst flow, and regarding the area where the node group with the largest water demand rise compared with historical data in the detection clusters is located as the pipe burst occurrence position.
After the pipe burst flow qualitative analysis in the step (41) passes, before the step (42) is executed, the method further comprises a step (41 a): if the outlier is not prompted again within T hours after the moment to be detected, the tube explosion continuity analysis is failed, and it is judged that tube explosion does not occur at the moment to be detected; if the outlier is prompted for the second time within T hours and the outlier passes the tube explosion flow qualitative analysis in the step (41), the tube explosion continuity analysis is judged to pass, the tube explosion occurs, the step (42) is entered,
further, the step (42) is specifically: and (4) regarding all detection clusters of which the moment to be detected is identified as an outlier and which pass through the pipe bursting flow qualitative analysis in the step (41) and the pipe bursting continuity analysis in the step (41a), regarding the area where the node group with the water demand rising most compared with the historical data in the detection clusters is located as the pipe bursting occurrence position.
The step (41a) is an optional step, although the false alarm rate can be effectively reduced, since the step needs to wait for a period of time, when the method is applied to the detection and positioning of the pipe burst, if the requirement on the detection speed is high, the step (41a) can be skipped, and manual recheck can be performed by combining other service systems; when the detection rate and the false alarm rate are required to be higher, the step (41a) can be used.
In this embodiment, a process of detecting and positioning a pipe burst of a water supply network from 1 to 2 points in the morning is taken as an example, and a process of detecting and positioning a pipe burst of a water supply network based on adaptive check is further described.
(1) Collecting pipe network monitoring data, and performing self-adaptive checking:
in the embodiment, the sample pipe network is adopted to simulate the data under the normal operation condition and the pipe explosion condition, the pipe explosion in the simulated data is detected and positioned by using the method, and the result is compared with the preset pipe explosion condition to judge the correctness of the result. The pipe network structure of the embodiment is shown in figure 2. The pipe network is supplied with water by two water sources, the pipe network comprises 3 high-level water pools, 91 nodes and 115 pipelines, the daily water supply amount is about 11 ten thousand tons, and the total length of the pipelines is 215.4 kilometers. 13 flow monitoring points and 14 pressure monitoring points are arranged in the pipeline network. Example setting n d15, extracting pressure and flow monitoring data at the moment to be detected and 15 days before the moment to be detected; the sampling time interval of the monitoring points is 15 minutes, so that the daily 24h/15min is 96 groups of monitoring data. The unknown pipeline roughness coefficients and node water demand in the pipe network are divided into 10 groups, as shown in fig. 2.
The location of the occurrence of a burst in the example is shown in fig. 2. The pipe explosion happens at 1-2 points in the morning, and the outlet flow rate during pipe explosion is about 100L/s; typical pressure and flow monitoring points on the day of pipe bursting are monitored in red boxes in fig. 2. All monitoring points are at the time to be detected and the history ndThe daily data is input into the adaptive check algorithm in real time, and the water demand check value of each node group at each historical moment and the moment to be detected can be obtained, as shown in fig. 3.
(2) Preprocessing a water demand checking result:
and defining the node groups with the upstream and downstream relations as the same detection cluster. The detection clusters in this case are respectively: e1＝{1,2,3}，E2＝{3,4,5}，E3＝{3,4,6}，E4＝{4,5,7}，E5＝{3,6,7}，E6＝{6,7,8}，E7＝{7,8,10}，E8{8,9,10 }. And rearranging the water demand checking results of all the node groups according to the detection cluster, and then carrying out normalization processing to obtain the matrix to be detected.
(3) Judging whether the moment to be detected is an outlier or not by adopting a DBSCAN algorithm:
and inputting the matrix to be detected into a DBSCAN algorithm, taking each row as a detection sample, analyzing the clustering relation among the samples and finding out an outlier. Typically, the time to be detected corresponds to the last row of the matrix to be detected. Therefore, if the last row is an outlier, the next step is entered for further analysis; otherwise, the tube explosion is not generated at the moment to be detected. In this embodiment, since the packet 6 is detonated at the time to be detected, the water demand checking result of the packet is greatly increased compared with the historical value, as shown in (b) in fig. 3, and thus the detection cluster E3＝{3,4,6}、E5＝{3,6,7}、E6The last row of the corresponding matrix to be detected of {6,7,8} is determined as an outlier.
(4) And continuously analyzing whether the outlier is a burst pipe or not based on the pipe network flow response characteristics under the burst pipe event, and outputting a burst pipe detection positioning result.
In the outlier, the water demand of the nodes of the group 6 is higher than the average value of the nodes at the same moment in historical data, so that the qualitative analysis of the pipe burst flow is passed; inputting the monitoring data of a moment after the moment to be detected into the step for analysis, finding that the data is still an outlier and can pass the step (41), so that the tube explosion continuity analysis passes, and outputting a result: a pipe burst occurs. Further analysis found that the node water demand for packet 6 rose to the maximum of all outliers, thus outputting the result: the location where the pipe burst occurs is the area corresponding to group 6. The result is consistent with the preset result, and the detection positioning result is judged to be correct.
The above embodiments are merely examples and do not limit the scope of the present invention. These embodiments may be implemented in other various manners, and various omissions, substitutions, and changes may be made without departing from the technical spirit of the present invention.

Claims (10)

1. A water supply network pipe burst detection and positioning method based on self-adaptive check is characterized by comprising the following steps:
(1) collecting pressure and flow monitoring data of a water supply network, and estimating water demand of each node by using a self-adaptive check algorithm;
(2) processing the water demand checking result to construct a detection cluster;
(3) judging whether each detection cluster is an outlier or not by adopting a DBSCAN algorithm;
(4) and analyzing the outlier based on the pipe network flow response characteristics during pipe explosion, outputting a pipe explosion detection result, and judging the position of the pipe explosion.
2. The water supply pipe network pipe burst detecting and positioning method based on self-adaptive checking as claimed in claim 1, wherein the step (1) is specifically as follows:
(11) acquiring pipe network monitoring data, at least including n before the moment to be detecteddMonitoring pressure and flow at each moment in the day;
(12) grouping pipelines and nodes in a water supply network respectively, and setting the roughness coefficients of all the pipelines in the same group to be the same and the daily water consumption change curves of all the nodes in the same group to be the same;
(13) and inputting the pressure and flow monitoring data and the grouping information into a self-adaptive checking algorithm, and calculating the water demand value of the node at each moment in real time.
3. The water supply pipe network pipe burst detecting and positioning method based on the adaptive check as claimed in claim 2, wherein the step (13) is specifically as follows:
(131) checking the node water demand at the current moment according to the rough coefficient at the previous moment, the pressure at the current moment and the flow monitoring data to obtain a reasoning observation value of the node water demand at the current moment;
(132) checking the pipeline rough coefficient at the current moment according to the inference observation value of the node water demand at the current moment, the pressure at the current moment and the flow monitoring data to obtain the inference observation value of the pipeline rough coefficient at the current moment;
(133) and calculating the optimal estimated values of the rough coefficient and the node water demand according to the rough coefficient and the node water demand at the previous moment, the rough coefficient and the node water demand at the current moment and the inference observation value of the node water demand by using a Kalman filter algorithm, and taking the optimal estimated values as final checking results of the rough coefficient and the node water demand at the current moment.
4. The water supply pipe network pipe burst detecting and positioning method based on self-adaptive checking as claimed in claim 1, wherein the step (2) is specifically as follows:
(21) defining adjacent nodes with upper and lower relations to be grouped into a detection cluster, wherein each detection cluster comprises p node groups, and further dividing the water supply network into a plurality of detection clusters;
(22) suppose the time to be detected is the nthdAt time t of day, any one detection cluster EkThe following matrixes to be detected are respectively constructed:
Figure FDA0002689773910000021
Figure FDA0002689773910000022
to detect clusters EkThe ith node is grouped into a node water demand checking column vector at the time t,
Figure FDA0002689773910000023
element (1) of
Figure FDA0002689773910000024
Indicating a detection cluster EkGrouping the ith node into the node water demand checking list result at the time t on the jth day, wherein i is 1, 2 … …, p, j is 1,2……，nd；
(23) for Qt_kEach column of
Figure FDA0002689773910000025
Order to
Figure FDA0002689773910000026
Wherein, muQIs a vector
Figure FDA0002689773910000027
Mean value of all elements in, σQIs a vector
Figure FDA0002689773910000028
Standard deviation of all elements in (a);
(24) rewriting the matrix to be detected as
Figure FDA0002689773910000029
5. The method for detecting and positioning the burst pipe of the water supply pipe network based on the adaptive check as claimed in claim 4, wherein the step (3) is specifically as follows:
(31) determining algorithm parameters and gamma, wherein the parameters are the neighborhood radius of the DBSCAN algorithm, and gamma is the minimum clustering example number of the DBSCAN algorithm;
(32) judging whether the moment to be detected is an outlier or not by using a DBSCAN algorithm;
(33) and (4) if the moment to be detected is the outlier, otherwise, judging that the tube explosion does not occur at the moment to be detected.
6. The method for detecting and locating the burst pipe of the water supply pipe network based on the adaptive check as claimed in claim 5, wherein γ in the step (31) is determined as: γ is 2 × p.
7. The method for detecting and locating the burst of water supply pipe network based on the adaptive check as claimed in claim 5, wherein the step (31) is determined by:
to-be-detected matrix NQt_kEach line of (a) is respectively regarded as one point in the space, the Euclidean distance between every two points is respectively calculated, and n is obtained in totald(nd-1)/2 pairs of Euclidean distances, and arranging all the solved Euclidean distances from large to small to obtain a vector
Figure FDA0002689773910000031
Selecting a distance threshold xi, and ordering:
Figure FDA0002689773910000032
wherein round is an operation function, which means rounding to find the whole,
Figure FDA0002689773910000033
representing the vector dpTo middle
Figure FDA0002689773910000034
Data of individual elements.
8. The method for detecting and positioning burst pipes of a water supply network based on adaptive check as claimed in claim 5, wherein the step (32) is specifically as follows:
(321) to-be-detected matrix NQt_kEach line of the cluster mark C is respectively regarded as a point in the space, one point is randomly selected as a current point, and the cluster mark C is made to be 1;
(322) all the points with Euclidean distance smaller than the current point are taken as the adjacent points of the current point, if the number of the adjacent points is less than gamma, the current point is marked as an outlier, and the step (324) is carried out; otherwise, marking the current point as a core point of the cluster C;
(323) repeating the step (322) for each adjacent point of the current point in sequence until no new adjacent point appears and all the discovered adjacent points belong to the cluster C;
(324) and selecting the next unmarked point as the current point, marking the clustering mark C as C +1, repeating the steps (322) and (323), and repeatedly executing the steps until all the points are marked.
9. The water supply pipe network pipe burst detecting and positioning method based on the adaptive check as claimed in claim 8, wherein the step (4) is specifically as follows:
(41) comparing the water demand checking result of each group in the outlier with the value at the same moment in the historical data, if the water demand of each group is smaller than the average value of the water demand at the same moment in the historical data, the qualitative analysis of the pipe bursting flow fails, and judging that the pipe bursting does not occur at the moment to be detected; otherwise, the pipe bursting flow qualitative analysis is passed, the pipe bursting at the time to be detected is judged, and the step (42) is executed;
(42) and regarding all detection clusters which are identified as outliers at the moment to be detected and pass through the qualitative analysis of the pipe burst flow, and regarding the area where the node group with the largest water demand rise compared with historical data in the detection clusters is located as the pipe burst occurrence position.
10. The method for detecting and positioning burst of water supply network based on adaptive check as claimed in claim 9, wherein the qualitative analysis of the burst flow rate in step (41) after passing step (42) further comprises step (41 a): if the outlier is not prompted again within T hours after the moment to be detected, the tube explosion continuity analysis is failed, and it is judged that tube explosion does not occur at the moment to be detected; if the outlier is prompted for the second time within T hours and the outlier passes the tube explosion flow qualitative analysis in the step (41), the tube explosion continuity analysis is judged to pass, the tube explosion occurs, the step (42) is entered,
further, the step (42) is specifically: and (4) regarding all detection clusters of which the moment to be detected is identified as an outlier and which pass through the pipe bursting flow qualitative analysis in the step (41) and the pipe bursting continuity analysis in the step (41a), regarding the area where the node group with the water demand rising most compared with the historical data in the detection clusters is located as the pipe bursting occurrence position.
CN202010987621.9A 2020-09-18 2020-09-18 Water supply pipe network pipe burst detection and positioning method based on self-adaptive checking Active CN112097125B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202010987621.9A CN112097125B (en) 2020-09-18 2020-09-18 Water supply pipe network pipe burst detection and positioning method based on self-adaptive checking

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202010987621.9A CN112097125B (en) 2020-09-18 2020-09-18 Water supply pipe network pipe burst detection and positioning method based on self-adaptive checking

Publications (2)

Publication Number Publication Date
CN112097125A true CN112097125A (en) 2020-12-18
CN112097125B CN112097125B (en) 2021-07-20

Family

ID=73760486

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202010987621.9A Active CN112097125B (en) 2020-09-18 2020-09-18 Water supply pipe network pipe burst detection and positioning method based on self-adaptive checking

Country Status (1)

Country Link
CN (1) CN112097125B (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114857509A (en) * 2021-02-04 2022-08-05 西安普特流体控制有限公司 Pipe network pipe burst leakage monitoring method and device and platform positioning and verifying method

Citations (10)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN105930653A (en) * 2016-04-19 2016-09-07 清华大学 Pipe explosion pre-warning method based on metering zone flow monitoring data
CN106959608A (en) * 2017-02-27 2017-07-18 同济大学 A kind of water supply network seepage optimal control method based on cluster particle cluster algorithm
CN107869653A (en) * 2017-12-14 2018-04-03 北京化工大学 A kind of pipeline flow sensitivity matrix soil's rigidity method
CN108647470A (en) * 2018-05-29 2018-10-12 杭州电子科技大学 A kind of localization method at the beginning of based on the leakage loss with depth belief network is clustered
CN108799844A (en) * 2018-06-07 2018-11-13 哈尔滨学院 A kind of water supply network pressure monitoring point site selecting method based on fuzzy set
US20180341859A1 (en) * 2017-05-24 2018-11-29 Southwest Research Institute Detection of Hazardous Leaks from Pipelines Using Optical Imaging and Neural Network
CN110108328A (en) * 2019-04-03 2019-08-09 同济大学 A kind of acquisition methods of water supply network leakage loss region water leakage
CN110285330A (en) * 2019-07-11 2019-09-27 重庆大学 A kind of water utilities net pipe burst detection method based on the local factor that peels off
CN111006137A (en) * 2019-12-18 2020-04-14 北京无线电计量测试研究所 Water supply pipeline leakage monitoring and leakage positioning method and system
CN111578154A (en) * 2020-05-25 2020-08-25 吉林大学 LSDR-JMI-based water supply network multi-leakage pressure sensor optimal arrangement method

Patent Citations (10)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN105930653A (en) * 2016-04-19 2016-09-07 清华大学 Pipe explosion pre-warning method based on metering zone flow monitoring data
CN106959608A (en) * 2017-02-27 2017-07-18 同济大学 A kind of water supply network seepage optimal control method based on cluster particle cluster algorithm
US20180341859A1 (en) * 2017-05-24 2018-11-29 Southwest Research Institute Detection of Hazardous Leaks from Pipelines Using Optical Imaging and Neural Network
CN107869653A (en) * 2017-12-14 2018-04-03 北京化工大学 A kind of pipeline flow sensitivity matrix soil's rigidity method
CN108647470A (en) * 2018-05-29 2018-10-12 杭州电子科技大学 A kind of localization method at the beginning of based on the leakage loss with depth belief network is clustered
CN108799844A (en) * 2018-06-07 2018-11-13 哈尔滨学院 A kind of water supply network pressure monitoring point site selecting method based on fuzzy set
CN110108328A (en) * 2019-04-03 2019-08-09 同济大学 A kind of acquisition methods of water supply network leakage loss region water leakage
CN110285330A (en) * 2019-07-11 2019-09-27 重庆大学 A kind of water utilities net pipe burst detection method based on the local factor that peels off
CN111006137A (en) * 2019-12-18 2020-04-14 北京无线电计量测试研究所 Water supply pipeline leakage monitoring and leakage positioning method and system
CN111578154A (en) * 2020-05-25 2020-08-25 吉林大学 LSDR-JMI-based water supply network multi-leakage pressure sensor optimal arrangement method

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
马力辉: "系统聚类分析用于供水管道漏损控制的研究", 《中国给水排水》 *

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114857509A (en) * 2021-02-04 2022-08-05 西安普特流体控制有限公司 Pipe network pipe burst leakage monitoring method and device and platform positioning and verifying method
CN114857509B (en) * 2021-02-04 2024-06-07 西安普特流体控制有限公司 Pipe network pipe explosion leakage monitoring method and device and platform positioning and verifying method

Also Published As

Publication number Publication date
CN112097125B (en) 2021-07-20

Similar Documents

Publication Publication Date Title
CN108984873B (en) Water supply network real-time leakage detection method, device, system and storage medium
US11494853B2 (en) Method for acquiring water leakage amount of leakage area in water distribution system
CN102235575B (en) Data processing method and system for checking pipeline leakage
CN103939749B (en) Flow circuits based on big data leakage intelligent adaptive monitoring system and method
Laucelli et al. Detecting anomalies in water distribution networks using EPR modelling paradigm
CN107869653B (en) Pipeline flow sensitivity matrix leakage detection method
CN112097126B (en) Water supply network pipe burst pipeline accurate identification method based on deep neural network
CN109442221B (en) A kind of water supply network booster method for detecting extracted based on pressure disturbance
CN110929359B (en) Pipe network siltation risk prediction modeling method based on PNN neural network and SWMM technology
CN112836758B (en) Independent metering area pipe network leakage detection method based on double-layer circulating neural network
CN110285330B (en) Water network pipe burst detection method based on local outlier factor
CN112097125B (en) Water supply pipe network pipe burst detection and positioning method based on self-adaptive checking
CN117540329B (en) Online early warning method and system for defects of drainage pipe network based on machine learning
CN115270372A (en) Drainage pipe network siltation judgment method based on depth sequence model
CN113128709B (en) Water supply network leakage management system
CN112780953B (en) Independent metering area pipe network leakage detection method based on mode detection
CN116823067B (en) Method and device for determining water quality cleaning state of pipe network and electronic equipment
CN113486950A (en) Intelligent pipe network water leakage detection method and system
CN117520989A (en) Natural gas pipeline leakage detection method based on machine learning
CN116484219A (en) Water supply network water quality abnormal pollution source identification method based on gate control graph neural network
KR20180083480A (en) The method of water quality forecasting based on IoT (Internet of Things)
CN113970073B (en) ResNet-based water supply network leakage accurate positioning method
CN115493093A (en) Steam heating pipe network leakage positioning method and system based on mechanical simulation
Rosich et al. Optimal sensor placement for leakage detection and isolation in water distribution networks
Snider et al. A flexible, leak crew focused localization model using a maximum coverage search area algorithm

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant