CN113775939B - Online identification and positioning method for newly increased leakage of water supply pipe network - Google Patents

Abstract

The invention discloses an online identification and positioning method for newly added leakage of a water supply pipe network. Through the combination of a hydraulic model of the water supply pipe network and a pressure monitoring system, online identification and rapid positioning of newly added leakage events are realized. The method includes two parts: an offline module and an online module: in the offline module, the response threshold and level of the monitoring system to new leakage events are determined through the statistical analysis of the historical data of the monitoring points, and the hydraulic model of the pipeline network is used to preview the hydraulic pressure of the new leakage. Build a database of new leakage alarm response modes in the pipeline network; the online module realizes online identification and rapid positioning of new leakage events by performing over-threshold discrimination and pattern matching on the real-time monitoring data of pressure monitoring points. The method of the invention has the characteristics of high leakage identification and positioning accuracy, high speed, and easy implementation, and can effectively improve the efficiency of water supply pipe network leakage management and control.

Description

An online identification and location method for new leakage in water supply network

technical field

The invention belongs to the technical field of urban water supply pipe networks, and relates to leakage monitoring and positioning of water supply pipe networks, in particular to an online identification and positioning method for newly added leakage of water supply pipe networks.

Background technique

As an important part of urban infrastructure, the water supply network system is an important guarantee for my country's urban economic development and residents' lives. However, many urban water supply systems in my country still have high leakage rates, and pipe bursts occur from time to time, resulting in huge waste of water resources and social and economic losses, and have seriously threatened urban public safety. Therefore, the leakage management of water supply enterprises has always been a hot and difficult point in the water supply industry, whether it is from the perspective of water resource protection or the guarantee of economic and social benefits of water supply enterprises.

At present, the leakage monitoring technology of water supply network can be roughly divided into four types: area-based metering, model-based, data-driven and device-based, or a mixture of these methods. Among the many existing detection methods for pipeline network leakage, most of them are difficult to achieve ideal results in practical applications due to their shortcomings such as complex technical principles, harsh operating conditions, inconvenient operation or high leak detection results. At this stage, the problem of accurate response and precise positioning of leakage events in the water supply network has not been effectively solved, and the phenomenon of passive leak detection and "looking at the leak and sighing" is still common.

Therefore, based on the current situation and needs of leakage management of my country's urban water supply network, combined with information and intelligent management methods of water supply network, a more effective and practical technical method for water supply network leakage identification and location is established to realize Efficient and safe water supply is still one of the important strategic development goals of the water supply industry.

Contents of the invention

Aiming at the deficiencies of the prior art, the present invention provides an online identification and location method for newly added leakage in the water supply pipe network. By combining the hydraulic model of the water supply pipe network with the monitoring system, the online detection of newly added leakage events in the water supply pipe network is realized. Identification and rapid localization of leaky areas.

The present invention is achieved through the following technical solutions:

An online identification and location method for newly added leakage in a water supply pipe network, which is realized by two parts: an offline module and an online module, wherein the offline module uses a hydraulic model of the water supply pipe network to preview the hydraulic state of the pipe network under the situation of newly added leakage Responding to the situation, constructing a new leakage alarm response mode database of the water supply pipe network; the online module discriminates and matches the patterns of the real-time monitoring data of the pressure monitoring points in the pipe network, so as to realize the online identification and rapid positioning of the newly added leakage event; Specific steps are as follows:

The operation of the offline module includes the following steps (1) to (5):

(1) Based on the historical monitoring data of pressure monitoring points in the pipeline network, determine the leakage response threshold of each pressure monitoring point at each moment:

C(m, t) = P _x ([P ₁ (t), P ₂ (t), ..., P _N (t)] ^T ) formula I;

In formula I: C(m, t) is the leakage response threshold of pressure monitoring equipment m at time t, m=1,..., M, M is the total number of pressure monitoring equipment in the pipeline network; P _x (*) is the lower limit of the xth percentile of the historical data sequence of the monitoring point, indicating that when the pressure of the monitoring point is lower than the leakage response threshold, a new leakage or pipe burst event may have occurred in the pipeline network; [P ₁ (t) , P ₂ (t),..., P _N (t)] ^T is the pressure monitoring historical data sequence of the pressure monitoring equipment m at the same time t every day, and N represents the number of historical monitoring values;

(2) Set the alarm level of each monitoring point to indicate the extent to which the observed pressure value of the monitoring point is less than the leakage response threshold when an abnormal event occurs. The alarm level L(m, t) of the monitoring device m at time t can be expressed as:

In formula II: h(m, t) is the observed value of pressure monitoring equipment m at time t, C(m, t)-h(m, t) represents the difference between the leakage response threshold and the real observed value; k= 1,..., K is the alarm level, K is the highest alarm level; L _k is the boundary value of the pressure difference corresponding to the alarm level k, which is set according to the engineer's experience and application requirements;

(3) Build a 24-hour continuous micro-hydraulic model of the water supply network, and perform model parameter verification and model dynamic update maintenance to meet the accuracy requirements of the actual application of the hydraulic model;

(4) Based on the hydraulic model of the pipe network constructed in step (3), the pressure-driven water volume model is used to simulate new leakage events, and the leakage events at different times, different locations, and different leakage flow conditions are simulated sequentially, and a sample of leakage events is constructed library, including the following steps:

(4+1) Initialize the hydraulic model of the pipe network based on the pressure-driven water volume, set the simulation time step Δt, the total duration of the simulation T, the minimum value of water loss q _min and the maximum value q _max , and the incremental step size of water loss Δq;

(4-2) According to the simulation time t and leakage Q _leak , traverse all nodes J _i , use the pressure-driven water quantity model to simulate the hydraulic state of the pipe network under the leakage event, record the generated leakage The sequential storage of water volume and pressure analog values of monitoring points is as follows:

Any leakage event: {time t, leakage node J _i , leakage quantity Q _leak , pressure analog values of all monitoring points: [monitoring point 1 pressure analog value, ..., monitoring point M pressure analog value]};

Among them, t=1,...,T; Q _leak =q _min : Δq:q _max ; J _i =J ₁ ,...,J _I , I is the total number of nodes in the pipe network;

(4-3) According to formula I and formula II, determine the alarm level of each monitoring point caused by each leakage event, constitute the alarm response mode of the monitoring point, and eliminate the leakage event without any alarm, as follows:

Any leakage event: {time t, leakage node J _i , leakage quantity Q _leak , monitoring point alarm response mode: [monitoring point 1 alarm level, ..., monitoring point M alarm level]};

Among them, t=1,...,T; Q _leak =q _min : Δq:q _max ; J _i =J ₁ ,...,J _I , I is the total number of nodes in the pipe network;

And thus build a sample library of water supply network leakage events;

(5) Perform clustering processing on the constructed water supply pipeline network leakage event sample library to further construct a new pipeline network leakage alarm response model database, which specifically includes the following steps:

(5+1) Classify the leakage events at the same time, at the same leakage node, with different leakage amounts, but with the same alarm response mode at the pressure monitoring point, and keep the smallest leakage amount to indicate that the alarm response mode occurs at this time It may be caused by the occurrence of the minimum water loss at the node;

(5-2) Further, the leakage events at the same time, different leakage nodes, but with the same alarm response mode at the pressure monitoring point are classified into one category, indicating that the possible leakage event corresponding to the alarm response mode occurs at this time, Nodes where leakage occurs and the minimum amount of water leakage can be expressed as:

{time t, monitoring point alarm response mode: [monitoring point 1 alarm level, ..., monitoring point M alarm level], the set of possible leakage nodes, the minimum water loss set corresponding to the possible leakage nodes};

Among them, t=1,...,T;

The operation of the online module includes the following steps (6) to (8):

(6) Obtain the monitoring data of each pressure monitoring point in real time through the online monitoring system, and judge whether the real-time monitoring data triggers a system alarm: when the monitoring data of at least one pressure monitoring point is lower than the current leakage response threshold, the water supply network Generate a new leakage event alarm, realize online identification, and continue to step (7); when the system is in normal operation, there is no need to perform leakage event alarm and location, that is, terminate the operation;

(7) Determine the alarm level of each pressure monitoring point according to the online monitoring data and formula II of each pressure monitoring point, and obtain the alarm response mode of the pressure monitoring point at the current moment;

(8) Match the alarm response mode of the pressure monitoring point at the current moment with the newly added leakage alarm response mode database of the water supply network to obtain the possible leakage nodes and the minimum water loss corresponding to the most matching alarm response mode, that is, the current The possible leakage area in the case of a new leakage alarm, thereby realizing the rapid positioning of the new leakage;

(9) The water supply company can send leakage detection personnel to the located leakage area to further find the specific location of the leakage or take other leakage management measures.

Preferably, in the step (1), the percentile Px(*) of the monitoring point historical data sequence is set to: x=5%～10%, which can be dynamically adjusted according to actual application requirements and effects, as follows: The larger the x value is, the larger the leakage response threshold is, the more sensitive the response to the leakage event is, but at the same time, a higher false alarm rate may be generated; on the contrary, the smaller the x value is, the smaller the leakage response threshold is, and the response to the leakage event is more sensitive. The less sensitive the incident response, the less likely it is to respond to real emergencies in a timely manner.

Preferably, in the step (2), the boundary value of the pressure difference can be set according to the experience of the engineer, specifically, it can be set as: L ₁ =0.2m, L ₂ =0.4m, L ₃ =0.6m, L ₄ = 0.8m, L ₅ =1.0m, corresponding to alarm level 1~5 respectively.

Preferably, in the step (3), different types of hydraulic models of the pipe network can be constructed according to the actual operating status of the water supply pipe network, and applied in subsequent steps, as follows: When there are obvious differences in the operating status, two types of pipe network hydraulic models are constructed on weekdays and rest days, and in the subsequent steps, the newly added leakage alarm response model databases on weekdays and rest days are respectively constructed for application.

Preferably, in the step (3), the hydraulic model of the water supply pipe network should be dynamically maintained and updated to ensure that the simulation results of the pipe network model and the hydraulic state of the real pipe network are updated synchronously. The specific maintenance update period can be determined according to the simulation accuracy of the hydraulic model. Changes are confirmed, not less than once a quarter.

Preferably, in the step (4-1), the initialization parameters of the hydraulic model of the pipe network based on the pressure-driven water volume are set to: time step Δt=1h, total simulation duration T=24h, minimum value of water loss q _min ≥ 5m ³ /h and the maximum value q _max ≤ 100m ³ /h, the incremental step size of leakage water is Δq=5m ³ /h.

Preferably, in the step (6), when a new leakage event alarm occurs, it is possible to check in advance the operation of equipment such as water pumps, valves, and fire hydrants in the pipeline network to determine whether the abnormal alarm is caused by the action of these equipment, thereby Reduce the false positive rate of leakage events.

The beneficial effects of the present invention are as follows:

(1) The technical principle of the method of the present invention is clear and definite, that is, first preview the possible hydraulic state of leakage in the water supply pipe network through the offline module, and then match the real-time monitoring data response situation through the online module to realize the online identification and detection of newly added leakage Regional positioning, and this technology only involves simple statistical analysis of data, hydraulic simulation of pipe network and other technologies, which is easy for engineering personnel to master and apply. Therefore, the method of the present invention is simple in application and strong in operability.

(2) The method of the present invention follows the basic principle of hydraulics, and the hydraulic state response under the newly added leakage state in the water supply pipe network is previewed through offline hydraulic simulation analysis, and the accuracy of the leakage location result is high; the online module of the present invention only passes through the Simple judgment and pattern matching of online monitoring data can realize online identification and rapid location of newly added leakage events, with high application efficiency and meeting online and real-time requirements.

(3) The method of the present invention can usually reduce the leakage area to a range of 1-5 km, and when multiple pressure monitoring points alarm at the same time, the leakage area can be further reduced, which greatly improves the efficiency of leakage location.

Description of drawings

Fig. 1 is the flow chart of the online identification and location method of newly-increased leakage of water supply pipe network of the present invention;

Fig. 2 is the schematic diagram of the topological structure of the water supply pipe network and the distribution diagram of the pressure monitoring points of embodiment 2;

Fig. 3 is the newly-increased leakage event and its positioning result of the monitoring point _M6 monitored by the method of the present invention in embodiment 2: (a) is an online recognition diagram of a leakage event; (b) is a location map of a leakage area;

Fig. 4 is the location result of leakage area when multiple pressure monitoring points alarm at the same time in embodiment 2.

detailed description

In order to make the method of the present invention easy to understand, the technical solution of the present invention will be clearly and completely described below in conjunction with the accompanying drawings and specific embodiments. Apparently, the described embodiments are only some of the embodiments of the application, not all of them. Based on the embodiments in this application, all other embodiments obtained by persons of ordinary skill in the art without making creative efforts belong to the scope of protection of this application.

Example 1

An online identification and location method for newly added leakage in a water supply network, as shown in Figure 1 is the technical roadmap for the implementation of this method. In this embodiment, a virtual water supply network including 3 online pressure monitoring points (marked as M ₁ -M ₃ ) and 10 nodes (marked as J ₁ -J ₁₀ ) is taken as an example. The specific steps are as follows:

This method is realized through two parts, an offline module and an online module.

First, execute the offline module in this method, and use the hydraulic model of the water supply network to preview the hydraulic state response of the pipeline network in the case of new leakage, so as to build a new leakage alarm response model database of the pipeline network, as follows:

(1) Collect the historical data of the pressure monitoring points of the water supply network, and determine the leakage response threshold of each pressure monitoring point at each moment according to formula I.

C(m, t) = P _x ([P ₁ (t), P ₂ (t), ..., P _N (t)] ^T ) formula I;

In formula I: C(m, t) is the leakage response threshold of pressure monitoring equipment m at time t, m=1,..., M, M is the total number of pressure monitoring equipment in the pipeline network; P _x (*) is the lower limit of the xth percentile of the historical data sequence of the monitoring point, indicating that when the pressure of the monitoring point is lower than the leakage response threshold, a new leakage or pipe burst event may have occurred in the pipeline network; [P ₁ (t) , P ₂ (t),..., P _N (t)] ^T is the pressure monitoring historical data sequence of the pressure monitoring device m at the same time t every day, and N represents the number of historical monitoring values.

The percentile P _x (*) of the historical data sequence of the monitoring point is usually set as: x=5%~10%, which can be dynamically adjusted according to the actual application requirements and effects. Generally speaking, the larger the value of x, the better the leakage response. The larger the threshold, the more sensitive the response to leakage events, but at the same time may produce a higher false alarm rate; on the contrary, the smaller the x value, the smaller the leakage response threshold, the less sensitive the response to leakage events, and may not be able to respond to leakage events. Respond promptly to real emergencies.

For the example pipe network, collect the historical monitoring data of 3 pressure monitoring points in the past 3 months, set the lower limit of percentile P _x (*) and take x=5%, and get the leakage of each pressure monitoring point at each moment Response thresholds are shown in Table 1 below.

Table 1 Leakage response threshold of each pressure monitoring point at each moment (5% lower limit of historical monitoring data)

(2) Set the alarm level of each monitoring point to indicate the extent to which the observed pressure value of the monitoring point is less than the leakage response threshold when an abnormal event occurs. The alarm level L(m, t) of the monitoring device m at time t can be expressed as:

In formula II: h(m, t) is the observed value of pressure monitoring equipment m at time t, C(m, t)-h(m, t) represents the difference between the leakage response threshold and the real observed value; k= 1,..., K is the alarm level, K is the highest alarm level; L _k is the boundary value of the pressure difference corresponding to the alarm level k, which is set according to the engineer's experience and application requirements.

The boundary value of the pressure difference can be set according to the experience of the engineer, specifically: L ₁ = 0.2m, L ₂ = 0.4m, L ₃ = 0.6m, L ₄ = 0.8m, L ₅ = 1.0m, corresponding to Alarm level 1~5.

Set the alarm level of each pressure monitoring point at each time, which can be set uniformly or separately.

For the sample pipe network, the alarm level of each pressure monitoring point at each time is uniformly set as follows (refer to formula II):

(3) Construct a 24-hour continuous micro-hydraulic model of the water supply network, and check the model parameters to meet the accuracy requirements of the actual application of the hydraulic model.

In order to further improve the accuracy of this method, different types of pipe network hydraulic models can be constructed according to the actual water supply network operating status and applied in subsequent steps. If there are obvious differences, two types of pipe network hydraulic models for working days and rest days are constructed respectively, and the new leakage alarm response mode databases for working days and rest days are respectively constructed in the subsequent steps for application.

In order to ensure the accuracy of this method in practical applications, the hydraulic model of the water supply network should be dynamically maintained and updated to ensure that the simulation results of the network model and the hydraulic state of the real network are updated synchronously. The specific maintenance update cycle can be determined according to the simulation accuracy of the hydraulic model. Changes are determined, usually not less than once a quarter

(4) Based on the hydraulic model of the pipe network constructed in step (3), the pressure-driven water volume model (PDD) is used to simulate new leakage events, and the leakage events at different times, different locations, and different leakage flow conditions are simulated in turn, and the construction Leakage event sample library.

Taking the sample pipe network as an example, the implementation steps include:

(4-1) Initialize the hydraulic model of the pipe network based on the pressure-driven water volume, set the simulation time step Δt = 1h, the total simulation duration T = 24h, the minimum value of leakage water q _min = 5m ³ /h and the maximum value q _max = 15m ³ /h, the incremental step size of leakage water loss is Δq=5m ³ /h.

(4-2) For the simulation time t (t=1,...,24), the leakage amount Q _leak (Q _leak =q _min :Δq:q _max =5:5:15m ³ /h), traverse all nodes J _i (i=1,...,10), use the PDD model to simulate the hydraulic state of the pipe network under the leakage event, record the generated leakage event results, and store them in the order of time, node, water loss, and pressure simulation value of the monitoring point, as shown in the following table 2.

Table 2 Hydraulic simulation results of the pipeline network under the new leakage event

(4-3) According to the alarm level of each pressure monitoring point determined in step (2) at each moment, determine the alarm level of each leakage event (constituting the alarm response mode of the monitoring point), and eliminate the leakage event without any alarm, In this way, a sample library of water supply network leakage events was constructed. Table 3 below shows the alarm levels of newly added leakage events in the example pipeline network.

Table 3 Alarm levels of newly added leakage events

Note in Table 3: The leakage event mark "#_#_#" means "time_leakage node_leakage volume".

It can be seen from Table 3 that when the leakage of node J1 is 5m ₃ /h, the event cannot be detected by the ^three pressure monitoring points, so the leakage event is eliminated, and finally the preprocessed leakage event sample library is obtained. As shown in Table 4 below.

Table 4 New Leakage Event Sample Library

Note in Table 4: The leakage event mark "#_#_#" means "time_leakage node_leakage volume".

(5) Perform clustering processing on the constructed water supply pipeline network leakage event sample database to further construct a new pipeline network leakage alarm response model database.

Taking the sample pipe network as an example, the implementation steps include:

(5-1) Classify the leakage events at the same time, at the same leakage node, with different leakage amounts, but with the same alarm response mode at the pressure monitoring point, into one category, and keep the smallest leakage amount. In Table 4, the alarm response modes of the monitoring points of the leakage events "1_J ₁₀ _5" and "1_J ₁₀ _10" are both [0,1,0], indicating that the system responds the same when these two leakage events occur, so it should be Classified into one category, the leakage event "1_J ₁₀ _5" with a smaller amount of leakage is retained, and the leakage event "1_J ₁₀ _10" is deleted. The results are shown in Table 5 below.

Table 5 New sample library of leakage events (after clustering)

Note in Table 5: The leakage event mark "#_#_#" means "time_leakage node_leakage amount".

(5-2) Further, the leakage events at the same time, different leakage nodes, but with the same alarm response mode at the pressure monitoring point are classified into one category, and a new leakage alarm response mode database for the pipeline network is constructed. In Table 5, when time t=1, the alarm response modes of the monitoring points of leakage events "1_J ₁ _10" and "1_J ₁₀ _5" are the same, and should be classified into one category. The results are shown in Table 6 below.

Table 6 New Leakage Alarm Response Model Database for Pipeline Network

Note in Table 6: "/..." indicates the leakage node with the same alarm response mode of other unlisted monitoring points.

Then, based on the newly added leakage alarm response mode database of the pipeline network constructed in steps (1) to (5), the online module in this method is run online to discriminate and match the real-time monitoring data of the pressure monitoring points in the pipeline network , in order to realize the online identification and rapid location of new leaks in the pipeline network, the details are as follows:

(6) For the current running time, obtain the monitoring data of each pressure monitoring point through the online monitoring system, and judge whether the real-time monitoring data triggers a system alarm, that is, if the monitoring data of at least one pressure monitoring point is lower than the leakage response at the current moment Threshold, the water supply network generates an alarm for a new leakage event, realizes online identification, and proceeds to step (7); otherwise, the system is operating normally, and there is no need to perform an alarm and location for a leakage event.

When a new leakage event alarm occurs, the operation of pumps, valves, fire hydrants and other equipment in the pipeline network can be checked in advance to determine whether the abnormal alarm is caused by the action of these devices, thereby reducing the false alarm rate of leakage events.

For the example pipeline network, assume that the real-time monitoring data of the pressure monitoring points M ₁ , M ₂ and M ₃ at 0:15 on a certain day (belonging to the time range of t=1) are [25.6, 31.9, 18.6]m in sequence. Among them, the monitoring data of monitoring point _M2 is lower than the current leakage response threshold of 32.1m, indicating that a new leakage event alarm has occurred in the pipeline network, so proceed to step (7).

(7) Determine the alarm level of each pressure monitoring point according to the online monitoring data of each pressure monitoring point and formula II, and obtain the alarm response mode of the pressure monitoring point at the current moment.

For the example pipeline network, the alarm response mode corresponding to the real-time monitoring data [25.6, 31.9, 18.6] of the three pressure monitoring points of the example pipeline network in step (6) is [0, 1, 0].

(8) Match the alarm response mode of the pressure monitoring point at the current moment with the newly added leakage alarm response mode database of the water supply network to obtain the possible leakage nodes and the minimum water loss corresponding to the most matching alarm response mode, that is, the current The possible leakage area in the case of a new leakage alarm is added, so as to realize the rapid positioning of the new leakage.

The alarm response mode of the pressure monitoring points in the example pipeline network is [0,1,0]. After matching with the newly added leakage alarm response mode database, the possible leakage node can be quickly located as J ₁ /J ₁₀ / ..., and the minimum water loss of these leakage nodes that cause the corresponding leakage alarm response mode. According to the spatial distribution of the located leakage nodes, the possible leakage areas can be divided.

(9) Based on this, the water supply company can send leakage detection personnel to the located leakage area to further find the specific location of the leakage or take other leakage management measures.

Example 2

The implementation steps and application effects of Embodiment 1 will be described below in conjunction with actual application scenarios.

Figure 2 is a schematic diagram of the topological structure of an urban water supply network, which includes 2 water source pools, 451 pipes, 300 nodes and 15 online pressure monitoring points (M ₁ ~M ₁₅ ). The local water company has built the hydraulic model of the pipeline network to guide the operation, scheduling and management of the pipeline network. With the help of the hydraulic model of the pipe network, the method of Embodiment 1 is used to monitor and locate the newly added leakage in the operation of the water supply pipe network in real time.

First, apply the off-line module described in Example 1 to determine the leakage response thresholds of 15 pressure monitoring points in 24 hours a day, and use the hydraulic model of the pipeline network to preview new leakage events, and build a pipeline network new leakage alarm response model database . Then, apply the online module described in embodiment 1 to discriminate the real-time data of the pressure monitoring point, when a leakage alarm event occurs, match the alarm response mode of the real-time monitoring data with the newly added leakage alarm response mode database of the pipeline network, The possible leakage node locations and the corresponding minimum water loss are obtained, thereby locating the possible areas of new leakage in the water supply network.

Fig. 3 shows a newly added leakage event (Fig. 3(a)) and its leakage area location result (Fig. 3(b)) detected by the method of Embodiment 1 in the daily operation of the water supply network. In Figure 3(a), the dotted line represents the leakage response threshold of monitoring point _M6 in 24 hours a day, and the solid line represents the online real-time monitoring data of monitoring point _M6 on May 21, 2021. It can be seen from Fig. 3(a) that the pressure of the monitoring point suddenly drops at 14:15, which is lower than the leakage response threshold of the monitoring point at this time period, which triggers the alarm of the online monitoring system and identifies the leakage in the pipeline network event. In addition, the occurrence of a leakage event can also be confirmed when the pressure monitoring value of the monitoring point is less than the leakage response threshold at several subsequent times, as shown in Figure 3(a). According to the monitoring data of monitoring point _M6 at the time of alarm and the monitoring data of other pressure monitoring points (no alarm), determine the alarm response mode of the pressure monitoring point, and match it with the newly added leakage alarm response mode database of the pipeline network, and further determine the The possible leakage nodes and the minimum leakage water volume are shown in Fig. 3(b). According to the distribution of possible leakage nodes, the possible leakage pipelines associated with these leakage nodes can be identified, and then the possible leakage area can be located, as shown in Figure 3(b) where the shadow coverage area (the total length of the pipeline is about 4.1km, Accounting for about 2.0% of the total pipe length of the pipe network). The determined minimum water loss at the leakage node can provide a certain reference for the assessment of the leakage.

After determining the possible leakage area, the water company dispatched leak detection staff to locate the leak precisely near the area, and confirmed that the leak was located in the located area (as shown in Figure 3(b)). According to the minimum leakage water volume of leakage nodes around the leakage point, it can be estimated that the leakage water volume at the leakage point is roughly 5-10m ³ /h. Therefore, the application of Embodiment 2 shows that the method of Embodiment 1 significantly improves the efficiency of leakage monitoring and location, and ensures the accuracy of leakage location.

In addition, Fig. 4 shows another situation where multiple pressure monitoring points alarm at the same time, that is: when multiple pressure monitoring points (M ₅ , M ₇ and M ₈ ) alarm at the same time, the method of Embodiment 1 is used to locate the Possible leak area. It can be seen from Fig. 4 that since the real-time data information of three pressure monitoring points are used at the same time, the method of embodiment 1 can locate the leakage to a very small area (the total length of the pipeline is about 1.5km). This greatly improves the efficiency of leak location.

Claims (6)

1. An online identification and location method for newly-increased leakage in a water supply pipe network, characterized in that it is realized in two parts, an offline module and an online module, wherein the offline module uses a hydraulic model of the water supply pipe network to preview the newly-increased leakage situation According to the hydraulic state response of the pipe network, a new leakage alarm response mode database of the water supply pipe network is constructed; the online module discriminates and matches the real-time monitoring data of the pressure monitoring points in the pipe network to realize the detection of newly added leakage events. Online identification and rapid positioning; the specific steps are as follows: The operation of the offline module includes the following steps (1) to (5): (1) Based on the historical monitoring data of pressure monitoring points in the pipeline network, determine the leakage response threshold of each pressure monitoring point at each moment: C(m,t)＝P _x ([P ₁ (t), P ₂ (t),...,P _N (t)] ^T ) formula Ⅰ; In formula Ⅰ: C(m,t) is the leakage response threshold of pressure monitoring equipment m at time t, m=1,...,M, M is the total number of pressure monitoring equipment in the pipeline network; P _x (*) is the monitoring The lower limit of the x-th percentile of the historical point data sequence indicates that when the pressure at the monitoring point is lower than the leakage response threshold, a new leakage or burst event may have occurred in the pipeline network; [P ₁ (t),P ₂ (t),...,P _N (t)] ^T is the pressure monitoring historical data sequence of the pressure monitoring equipment m at the same time t every day, and N represents the number of historical monitoring values; The percentile P _x (*) of the historical data sequence of the monitoring point is set as: x=5%~10%, which can be dynamically adjusted according to the actual application requirements and effects, as follows: the larger the value of x, the higher the leakage response threshold. The larger the value of x, the more sensitive the response to leakage events, but at the same time may produce a higher false alarm rate; on the contrary, the smaller the value of x, the smaller the leakage response threshold, the less sensitive the response to leakage events, and may not be able to respond to real Respond to emergencies in a timely manner; (2) Set the alarm level of each monitoring point to indicate the extent to which the observed pressure value of the monitoring point is less than the leakage response threshold when an abnormal event occurs. The alarm level L(m,t) of the monitoring device m at time t can be expressed as:

In formula II: h(m,t) is the observed value of pressure monitoring equipment m at time t, C(m,t)-h(m,t) represents the difference between the leakage response threshold and the real observed value; k= 1,..., K is the alarm level, K is the highest alarm level; L _k is the boundary value of the pressure difference corresponding to the alarm level k, which is set according to the engineer's experience and application requirements; (3) Build a 24-hour continuous micro-hydraulic model of the water supply network, and perform model parameter verification and model dynamic update maintenance to meet the accuracy requirements of the actual application of the hydraulic model; (4) Based on the hydraulic model of the pipe network constructed in step (3), the pressure-driven water volume model is used to simulate new leakage events, and the leakage events at different times, different locations, and different leakage flow conditions are simulated sequentially, and a sample of leakage events is constructed library, including the following steps: (4-1) Initialize the hydraulic model of the pipe network based on the pressure-driven water volume, set the simulation time step Δt, the total duration of the simulation T, the minimum value q _min and the maximum value q _max of the leakage water, and the incremental step of the leakage water Δq; (4-2) According to the simulation time t and leakage Q _leak , traverse all nodes J _i , use the pressure-driven water quantity model to simulate the hydraulic state of the pipe network under the leakage event, record the generated leakage The sequential storage of water volume and pressure analog values of monitoring points is as follows: Any leakage event: {time t, leakage node J _i , leakage quantity Q _leak , pressure analog values of all monitoring points: [monitoring point 1 pressure analog value, ..., monitoring point M pressure analog value]}; Among them, t=1,...,T; Q _leak ＝q _min :Δq:q _max ; J _i ＝J ₁ ,...,J _I , I is the total number of nodes in the pipe network; (4-3) According to formula Ⅰ and formula Ⅱ, determine the alarm level of each monitoring point caused by each leakage event, constitute the alarm response mode of the monitoring point, and eliminate the leakage event without any alarm, as follows: Any leakage event: {time t, leakage node J _i , leakage quantity Q _leak , monitoring point alarm response mode: [monitoring point 1 alarm level, ..., monitoring point M alarm level]}; Among them, t=1,...,T; Q _leak ＝q _min :Δq:q _max ; J _i ＝J ₁ ,...,J _I , I is the total number of nodes in the pipe network; and thus construct the leakage event of the water supply pipe network sample library; (5) Perform clustering processing on the constructed water supply pipeline network leakage event sample library to further construct a new pipeline network leakage alarm response model database, which specifically includes the following steps: (5-1) Classify the leakage events at the same time, at the same leakage node, with different leakage amounts, but with the same alarm response mode at the pressure monitoring point, and keep the smallest leakage amount to indicate that the alarm response mode occurs at this time It may be caused by the occurrence of the minimum water loss at the node; (5-2) Further, the leakage events at the same time, different leakage nodes, but with the same alarm response mode at the pressure monitoring point are classified into one category, indicating that the possible leakage event corresponding to the alarm response mode occurs at this time, Nodes where leakage occurs and the minimum amount of water leakage can be expressed as: {time t, monitoring point alarm response mode: [monitoring point 1 alarm level, ..., monitoring point M alarm level], the set of possible leakage nodes, the minimum water loss set corresponding to the possible leakage nodes}; Among them, t=1,...,T; The operation of the online module includes the following steps (6) to (8): (6) Obtain the monitoring data of each pressure monitoring point in real time through the online monitoring system, and judge whether the real-time monitoring data triggers a system alarm: when the monitoring data of at least one pressure monitoring point is lower than the current leakage response threshold, the water supply network Generate a new leakage event alarm, realize online identification, and continue to step (7); when the system is in normal operation, there is no need to perform leakage event alarm and location, that is, terminate the operation; (7) Determine the alarm level of each pressure monitoring point according to the online monitoring data of each pressure monitoring point and formula II, and obtain the alarm response mode of the pressure monitoring point at the current moment; (8) Match the alarm response mode of the pressure monitoring point at the current moment with the newly added leakage alarm response mode database of the water supply network to obtain the possible leakage nodes and the minimum water loss corresponding to the most matching alarm response mode, that is, the current The possible leakage area in the case of a new leakage alarm, thereby realizing the rapid positioning of the new leakage; (9) The water supply company can send leakage detection personnel to the located leakage area to further find the specific location of the leakage or take other leakage management measures. 2. The online identification and location method of a new leakage in a water supply pipe network according to claim 1, characterized in that, in the step (2), the boundary value of the pressure difference can be set according to the engineer's experience, Specifically, it can be set as: L ₁ =0.2m, L ₂ =0.4m, L ₃ =0.6m, L ₄ =0.8m, L ₅ =1.0m, corresponding to alarm levels 1-5 respectively. 3. The online identification and location method of newly added leakage in a water supply pipe network according to claim 1, characterized in that, in the step (3), different types of leaks can be constructed according to the actual water supply pipe network operating conditions. The hydraulic model of the pipe network will be applied in the subsequent steps, as follows: when there are obvious differences in the operation status of the water supply pipe network on working days and rest days, two types of pipe network hydraulic models on working days and rest days will be constructed respectively, and will be used in the follow-up In the step, the newly added leakage alarm response mode databases for weekdays and rest days are respectively constructed for application. 4. The online identification and location method of newly added leakage of a kind of water supply pipe network according to claim 1, it is characterized in that, in described step (3), the hydraulic model of water supply pipe network should carry out dynamic maintenance update, to ensure The simulation results of the pipe network model are updated synchronously with the hydraulic state of the real pipe network. The specific maintenance update cycle can be determined according to the change of the simulation accuracy of the hydraulic model, and should not be less than once a quarter. 5. The online identification and location method of a new leakage in a water supply pipe network according to claim 1, wherein in said step (4-1), the initialization of the hydraulic model of the pipe network based on pressure-driven water flow The parameters are set as: time step Δt=1h, total duration of simulation T=24h, minimum value of water loss q _min ≥5m ³ /h and maximum value q _max ≤100m ³ /h, increasing step size of water loss Δq=5m ³ /h. 6. The online identification and location method of newly-increased leakage in a water supply pipe network according to claim 1, characterized in that, in said step (6), when an alarm for newly-increased leakage occurs, it can be pre-passed Check the operation of pumps, valves, and fire hydrants in the pipeline network to determine whether the abnormal alarm is caused by the action of these devices, thereby reducing the false alarm rate of leakage events.

Images