CN110837933A - Leakage identification method, device, equipment and storage medium based on neural network - Google Patents
Leakage identification method, device, equipment and storage medium based on neural network Download PDFInfo
- Publication number
- CN110837933A CN110837933A CN201911093678.8A CN201911093678A CN110837933A CN 110837933 A CN110837933 A CN 110837933A CN 201911093678 A CN201911093678 A CN 201911093678A CN 110837933 A CN110837933 A CN 110837933A
- Authority
- CN
- China
- Prior art keywords
- neural network
- value
- water consumption
- module
- time period
- 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.)
- Pending
Links
- 238000013528 artificial neural network Methods 0.000 title claims abstract description 93
- 238000000034 method Methods 0.000 title claims abstract description 48
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 claims abstract description 156
- 238000003062 neural network model Methods 0.000 claims abstract description 26
- 238000012545 processing Methods 0.000 claims abstract description 24
- 230000006870 function Effects 0.000 claims description 25
- 238000012549 training Methods 0.000 claims description 22
- 238000004590 computer program Methods 0.000 claims description 9
- 238000001556 precipitation Methods 0.000 claims description 4
- 238000004458 analytical method Methods 0.000 description 8
- 238000010586 diagram Methods 0.000 description 6
- 230000000694 effects Effects 0.000 description 4
- 239000002699 waste material Substances 0.000 description 3
- 238000004422 calculation algorithm Methods 0.000 description 2
- 238000001514 detection method Methods 0.000 description 2
- 238000010801 machine learning Methods 0.000 description 2
- 238000004519 manufacturing process Methods 0.000 description 2
- 238000013507 mapping Methods 0.000 description 2
- 238000012544 monitoring process Methods 0.000 description 2
- 230000002159 abnormal effect Effects 0.000 description 1
- 230000006978 adaptation Effects 0.000 description 1
- 230000003044 adaptive effect Effects 0.000 description 1
- 230000004075 alteration Effects 0.000 description 1
- 230000006399 behavior Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 230000007812 deficiency Effects 0.000 description 1
- 230000003111 delayed effect Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000011156 evaluation Methods 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 238000011835 investigation Methods 0.000 description 1
- 238000012423 maintenance Methods 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 238000005192 partition Methods 0.000 description 1
- 230000008447 perception Effects 0.000 description 1
- 230000002085 persistent effect Effects 0.000 description 1
- 230000008569 process Effects 0.000 description 1
- 238000011084 recovery Methods 0.000 description 1
- 238000005070 sampling Methods 0.000 description 1
- 230000004083 survival effect Effects 0.000 description 1
- 230000001360 synchronised effect Effects 0.000 description 1
- 238000012360 testing method Methods 0.000 description 1
- 230000001052 transient effect Effects 0.000 description 1
- 238000005406 washing Methods 0.000 description 1
Images
Classifications
-
- 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
- G06Q10/00—Administration; Management
- G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
- G06F16/29—Geographical information databases
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- 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)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Business, Economics & Management (AREA)
- General Physics & Mathematics (AREA)
- Health & Medical Sciences (AREA)
- Economics (AREA)
- Human Resources & Organizations (AREA)
- General Health & Medical Sciences (AREA)
- Data Mining & Analysis (AREA)
- General Engineering & Computer Science (AREA)
- Strategic Management (AREA)
- Tourism & Hospitality (AREA)
- Molecular Biology (AREA)
- Software Systems (AREA)
- Marketing (AREA)
- Mathematical Physics (AREA)
- Computing Systems (AREA)
- Databases & Information Systems (AREA)
- General Business, Economics & Management (AREA)
- Evolutionary Computation (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- Remote Sensing (AREA)
- Development Economics (AREA)
- Game Theory and Decision Science (AREA)
- Quality & Reliability (AREA)
- Operations Research (AREA)
- Entrepreneurship & Innovation (AREA)
- Public Health (AREA)
- Water Supply & Treatment (AREA)
- Primary Health Care (AREA)
- Examining Or Testing Airtightness (AREA)
Abstract
The invention discloses a leakage identification method, device, equipment and storage medium based on a neural network. The leakage identification method based on the neural network comprises the following steps: s1, acquiring flow meter data, meteorological information, a day-of-year value and a minute-of-day value; s2, calculating model input information through the data acquired in the step S1, wherein the model input information comprises water consumption of a time period in the region, a meteorological function, a time period function and a working day function; s3, processing the model input information by using the neural network model to obtain the estimated value of water consumption in the next time period region; s4, calculating a comparison value according to the estimated value of the water consumption of the region in the next time period and the actual water consumption of the region, and executing the step S5 if the comparison value is larger than a preset threshold value; when the comparison value is less than or equal to the preset threshold value, returning to the step S1; and S5, executing alarm operation when the comparison value is larger than the preset threshold value, and then returning to the step S1.
Description
Technical Field
The invention relates to the technical field of machine learning, in particular to a leakage identification method, a device, equipment and a storage medium based on a neural network.
Background
The existing water supply enterprises passively inspect leakage, and problems can be found only when the leakage time is long and the total water loss is large, so that the leakage position is located for maintenance, and the waste of water resources is serious; even if a leak detection company is requested to perform water leakage general investigation, a good effect can be obtained in a short time, but the purpose of reducing leakage cannot be fundamentally achieved due to the existence of the water leakage recovery phenomenon.
Leakage is one of the biggest pain points in the Water industry in China, and currently, an analysis mode is mainly managed through DMA (direct memory access independent Metering Area), and active leakage evaluation analysis and control are performed by adopting an IWA (IWA-International Water Association) Water supply system Water quantity balance table. The most important part in the water balance meter of the water supply system, namely the quantifiable water quantity (the charged metered water quantity and the uncharged metered water quantity), is acquired mainly by adopting a subarea flowmeter and a user water meter; the user water meters adopt a large number of mechanical meters, so that the popularization rate of the intelligent water meters is low; even if an intelligent water meter is adopted, the water quantity acquisition density of the intelligent water meter is low, and the meter reading cannot be synchronous; therefore, the balance table analysis method is difficult to implement accurately, the dynamic analysis performance is extremely poor, and the abnormal water amount cannot be found in time. In view of the above reasons, domestic attempts have been made to use analysis such as pipe network GIS (geographic information system) and dynamic hydraulic model modeling as a basis, apply various leakage calculation methods, such as night minimum flow/day average flow, night net flow per unit pipe length, night net flow per unit service connection, pipe network leakage Index (iliifradructureleakage Index), minimum night flow threshold alarm and leakage detection triggering, to manage metering partitions and DMAs, and to meter the amount of leakage water in regions and the amount of poor water produced and sold in regions. The current situations of leakage and poor production and sale of the pipe network and each stage of subareas are comprehensively analyzed according to different spatial scales and different time scales, and main influence factors of the leakage and poor production and sale are identified. However, in the dynamic hydraulic modeling, due to the fact that the actual condition of a pipeline and the dynamic of a user cannot be measured, the model is difficult to establish, and the model prediction effect is poor. The analysis period such as minimum flow is long, and the daytime anomaly is difficult to find.
Disclosure of Invention
The present invention is directed to overcome the above-mentioned deficiencies in the prior art, and provides a method, an apparatus, a device and a storage medium for leakage identification based on a neural network.
In order to achieve the above purpose, the invention provides the following technical scheme:
a leakage identification method based on a neural network comprises the following steps:
s1, acquiring flow meter data, meteorological information, a day-of-year value and a minute-of-day value;
s2, calculating model input information through the data acquired in the step S1, wherein the model input information comprises water consumption of a time period in the region, a meteorological function, a time period function and a working day function;
s3, processing the model input information by using the neural network model to obtain the estimated value of water consumption in the next time period region;
s4, calculating a comparison value according to the estimated value of the water consumption of the region in the next time period and the actual water consumption of the region, and executing the step S5 if the comparison value is larger than a preset threshold value; when the comparison value is less than or equal to the preset threshold value, returning to the step S1;
and S5, executing alarm operation when the comparison value is larger than the preset threshold value, and then returning to the step S1.
Preferably, the weather information acquired in step S1 includes temperature, humidity or precipitation information;
in the step S2, the meteorological function represents meteorological information, including temperature values or humidity values, and is represented in the form of an array; the time period function represents an annual date time interval and an intra-day time interval and is obtained by calculating an annual date value and an intra-day minute value respectively; the working day function represents working day or holiday information and is represented by the ratio of the time length of the date to the next nearest holiday to the end time length of the next nearest holiday.
Preferably, the neural network model is a BP neural network model.
Preferably, the training method of the neural network model comprises: and (4) extracting historical acquisition data by adopting a moving window method to train the neural network.
Preferably, the step S5 specifically includes the following steps:
s51, when the comparison value is larger than the preset threshold value, starting timing operation;
s52, in the timing time interval, if the time interval in which the comparison value exceeds the preset threshold value is larger than the preset time interval value, executing the step S53; otherwise, returning to the step S1;
and S53, executing alarm operation.
Preferably, the value of the comparison value exceeding the preset threshold is in inverse time function relation with the value of the preset time interval.
A leakage identification device based on a neural network comprises a data acquisition module, a database, a neural network identification module, a data processing module and an alarm module;
the output end of the data acquisition module is connected with a database, the database is respectively connected with the first input ends of the neural network identification module and the data processing module, the output end of the neural network identification module is connected with the second input end of the data processing module, and the output end of the data processing module is connected with the input end of the alarm module;
the data acquisition module is used for acquiring flowmeter data, meteorological information, an annual date value and an intraday minute value; the database is used for storing the data acquired by the data acquisition module; the neural network identification module is used for acquiring a water consumption estimation value of a region in the next time period according to related data of the region in the previous time period input by the database; the data processing module is used for calculating a comparison value according to the estimated value of the water consumption of the region in the next time period and the actual water consumption of the region, and judging whether to control the alarm module to execute alarm operation or not; the alarm module is used for executing alarm operation according to the output signal of the data processing module.
Preferably, the neural network-based leakage recognition apparatus further includes a neural network training module; the output end of the database is connected with the input end of the neural network training module, and the output end of the neural network training module is connected with the second input end of the neural network identification module; the neural network training module adopts a moving window method to extract historical acquisition data in a database to train the neural network.
A neural network-based leak identification apparatus, comprising:
a memory for storing a computer program;
a processor for implementing the steps of the neural network based leakage identification method when executing the computer program.
A readable storage medium, having stored thereon a computer program which, when being executed by a processor, carries out the steps of the neural network-based leakage identification method.
Compared with the prior art, the invention has the beneficial effects that:
1. the influence of factors such as daily water consumption, gas image, current time period, working day, holidays and the like of a user on the water consumption is comprehensively considered, the estimated value of the water consumption in the region of the next time period is more accurately predicted, the leakage phenomenon is more quickly and accurately identified, and the waste condition of water resources is reduced;
2. the neural network is trained by adopting a moving window method so as to acquire a recent newer data training model, the parameters of the neural network model are updated in time, and the prediction accuracy of the neural network model is enhanced;
3. when the leakage identification is carried out, the time interval when the comparison value exceeds the preset threshold value is larger than the preset time interval, and then the alarm operation is carried out, so that the influence of short-time accidental water consumption on the leakage identification is avoided; and meanwhile, the numerical value of the comparison value exceeding the preset threshold value and the preset time interval value are set to form an inverse time limit function relationship, so that the speed of leakage identification is increased.
Description of the drawings:
fig. 1 is a flowchart of a neural network-based leak identification method according to exemplary embodiment 1 of the present invention;
FIG. 2 is a diagram illustrating DMA management analysis patterns in accordance with exemplary embodiment 1 of the present invention;
FIG. 3 is a schematic diagram of a neural network recognition and training process in accordance with exemplary embodiment 1 of the present invention;
fig. 4 is a detailed flowchart of a neural network-based leak recognition method according to exemplary embodiment 1 of the present invention;
FIG. 5 is a graph showing the effect of a regional water usage estimate curve in exemplary embodiment 2 of the present invention; (a) comparing a tuesday actual water usage curve with a tuesday three-purpose water usage estimated value curve predicted according to a tuesday water usage, (b) comparing a tuesday actual water usage curve with a tuesday water usage estimated value curve predicted according to a tuesday water usage, (c) comparing a tuesday actual water usage curve with a tuesday water usage estimated value curve predicted according to a tuesday water usage, (d) comparing a sunday actual water usage curve with a tuesday water usage estimated value curve predicted according to a tuesday water usage, (e) comparing a tuesday three-purpose water usage curve with a tuesday three-purpose water usage estimated value curve predicted according to a tuesday water usage, (f) comparing a tuesday three-purpose water usage curve with a tuesday water usage estimated value curve predicted according to a tuesday water usage, (g) comparing a tuesday actual water usage curve with a tuesday three-purpose water usage estimated value curve predicted according to a tues, (h) comparing the actual water consumption curve of the sunday with the estimated value curve of the sunday water consumption predicted according to the water consumption of saturday;
fig. 6 is a schematic structural diagram of a leakage identification apparatus based on a neural network according to exemplary embodiment 3 of the present invention;
fig. 7 is a schematic structural diagram of a leakage identification device based on a neural network in exemplary embodiment 4 of the present invention;
fig. 8 is a schematic structural diagram of a leakage identification device based on a neural network in exemplary embodiment 4 of the present invention.
Detailed Description
The present invention will be described in further detail with reference to test examples and specific embodiments. It should be understood that the scope of the above-described subject matter is not limited to the following examples, and any techniques implemented based on the disclosure of the present invention are within the scope of the present invention.
Example 1
As shown in fig. 1, the present embodiment provides a leakage identification method based on a neural network, including the following steps:
s1, acquiring flow meter data, meteorological information, a day-of-year value and a minute-of-day value;
s2, calculating model input information through the data acquired in the step S1, wherein the model input information comprises water consumption of a time period in the region, a meteorological function, a time period function and a working day function;
s3, processing the model input information by using the neural network model to obtain the estimated value of water consumption in the next time period region;
s4, calculating a comparison value according to the estimated value of the water consumption of the region in the next time period and the actual water consumption of the region, and executing the step S5 if the comparison value is larger than a preset threshold value; when the comparison value is less than or equal to the preset threshold value, returning to the step S1;
and S5, executing alarm operation when the comparison value is larger than the preset threshold value, and then returning to the step S1.
A flowmeter is arranged on a water supply inlet pipe and a water supply outlet pipe of the monitoring area so as to accurately obtain the water consumption. The method comprises the steps of collecting data of each flowmeter in a monitoring area, and calculating water consumption in a time period in the area according to the data of each flowmeter. The water consumption of the last time period is combined with information such as meteorological information, a day value in the year, a minute value in the day and the like to be used as input information of the neural network model, and the water consumption estimation value of the area of the next time period is calculated through the neural network model. The estimated value comprehensively considers the influence of factors such as daily water consumption, gas image, current time period, working day, holidays and the like of a user on the water consumption. And then, calculating a comparison value according to the estimated value of the water consumption of the region and the actual water consumption of the region in the next time period, and executing alarm operation when the comparison value is greater than a preset threshold value. The method comprehensively considers the influence of various factors on the water consumption; and by utilizing the self-learning and self-adaptive capacity of the neural network, the water consumption estimation value of the next time period region can be predicted more accurately, the leakage phenomenon can be identified more quickly and accurately, and the waste condition of water resources is reduced.
As shown in fig. 2, the present invention installs flow meters in all the pipe sections entering and exiting a supervision area based on the existing DMA management analysis mode, and designates the value of all the flow meters flowing into the area as positive flow and the value of all the flow meters flowing out of the area as negative flow; and then, the flow is accumulated for a fixed time period of all the flow meters, and the water consumption in the region in the time period (the water consumption in the region comprises the water leakage of the pipeline and the facility and the water consumption of the user) is calculated, namely the water consumption in the region in the time period in the step S2. The fixed time length adopted by the statistical flow meter is set according to specific conditions and can be water flow in a minute interval, water consumption in an hour interval or water consumption in an hour interval and the like.
The weather information acquired in step S1 in this embodiment includes information such as temperature or humidity; the meteorological function described in step S2 represents meteorological information, including temperature values or humidity values, which may be represented in the form of an array.
The time period function described in step S2 represents an intra-year date-time interval and an intra-day time interval, which are calculated from an intra-year date value and an intra-day minute value, respectively;
the workday function described in step S2 represents information such as workday or holiday, and may be represented by a ratio of the date to the time length of the next closest holiday to the time length of the end of the next closest holiday.
The influence of factors such as daily water consumption of a user, gas image, current time period, working day, holidays and the like on the water consumption is comprehensively considered, and the water consumption estimation value of the area in the next time period is more accurately predicted. Meteorological factors such as temperature, humidity, precipitation, etc. can affect the amount of water used by a user. When the weather temperature is higher, the water consumption is obviously increased; such as water intake, water usage for washing, etc.; the agricultural water consumption is closely related to meteorological factors such as temperature, humidity and precipitation; therefore, the weather information is input, and the prediction accuracy can be improved. The difference of water consumption in different time periods in one day is large, so that the current time period needs to be comprehensively considered when the water consumption in the next time period is predicted; for example, the difference between the water consumption in the peak water consumption period and the minimum flow rate period at night is large; therefore, the accuracy of prediction can be improved by inputting the current time period information. The working day and holiday factors can influence the water consumption users and the water consumption of the users; for example, when a user goes out, the water using behavior is delayed; when mechanisms such as schools and the like leave a vacation, the water consumption can be greatly reduced, so that the probability of leaving home and going out of a user is reflected through a working day function, or the water consumption is reduced; therefore, the working day and holiday factors can improve the accuracy of prediction.
The neural network model adopted in step S3 is an existing neural network model, and the estimated value of water consumption in the next time period is calculated according to the self-learning and self-adaptive capabilities of the neural network model.
Preferably, the neural network model is a bp (back propagation) neural network model.
The BP neural network is a multi-layer feedforward network trained according to an error inverse propagation algorithm, and is one of the most widely applied neural network models at present. The BP network can learn and store a large number of input-output pattern mappings without prior disclosure of mathematical equations describing such mappings. The BP neural network is a machine learning algorithm which is better adaptive to nonlinearity and has continuous perception. And adopting a BP neural network to accurately calculate and obtain the estimated value of the water consumption of the area in the next time period.
As shown in fig. 3, further, the training method of the neural network model includes: and extracting the stored historical acquisition data in the database by adopting a moving window method to train the neural network, so that the time period of the training data keeps a fixed distance from the current time. The water consumption is greatly influenced by time; in order to ensure that the model can adapt to the gradual development of water consumption in an interval more quickly, the problem that the accuracy of the model is low due to the fact that the interval time of historical data used for training is long, for example, the model is influenced by factors such as the survival rate is avoided; therefore, it is necessary to acquire new data to train the neural network model and adjust the parameters inside the neural network model. By adopting the moving window method to train the neural network, a recent newer data training model can be acquired, the parameters of the neural network model can be updated in time, the adaptation of the system to the new water use trend is improved, and the prediction accuracy of the neural network model is enhanced.
Further, the difference between the actual water usage of the area and the estimated water usage of the area in the next time period is recorded as a comparison value in step S4.
As shown in fig. 4, further, step S5 specifically includes the following steps:
s51, when the comparison value is larger than the preset threshold value, starting timing operation;
s52, in the timing time interval, if the time interval in which the comparison value exceeds the preset threshold value is larger than the preset time interval value, executing the step S53; otherwise, returning to the step S1;
and S53, executing alarm operation.
The short-time accidental water consumption can cause that the actual water consumption of the region is larger than the estimated value of the water consumption of the region in the next time period, so that the misjudgment of leakage identification is caused. When the comparison value is larger than the preset threshold value, timing operation is started, and when the comparison value is longer than the preset threshold value, alarm operation is executed, so that misjudgment of leakage identification caused by short-time accidental water consumption is avoided.
Further, the value of the comparison value exceeding the preset threshold value and the preset time interval value form an inverse time limit function relationship. Namely, the larger the value of the comparison value exceeding the preset threshold value is, the smaller the value of the preset time interval is. The simplest of the inverse time-limited functions is an inverse proportional function. The numerical value of the comparison value exceeding the preset threshold value and the preset time interval value form an inverse time limit function relationship, so that whether leakage occurs or not can be judged more accurately, and meanwhile, when the leakage condition is serious, the alarm operation can be quickly performed by quickly responding.
Example 2
In this embodiment, the leakage identification method based on the neural network described in embodiment 1 is used to estimate the water consumption in a certain area to determine whether leakage occurs. In the embodiment, before the neural network is adopted to predict the water consumption estimation value of the area, the water consumption of the users in the area in two weeks of history is adopted to train the BP neural network, so that the prediction accuracy of the BP neural network is improved. When the regional water consumption estimation value is predicted, the acquisition period is set to be 15 minutes, the water consumption in the previous time period is acquired every 15 minutes, the future water consumption is predicted, the obtained regional water consumption estimation value curve effect graph is shown in figure 5, the horizontal axis of the graph shows the number of sampling periods from the prediction starting point, and the vertical axis shows the hourly flow.
The curves of estimated water consumption for the regions indicated by the dotted lines in the graphs (a) to (d) in fig. 5 are curves of estimated water consumption for the region on wednesday, thursday, saturday, sunday, respectively, which are predicted based on the water consumption on tuesday; the actual water usage curves indicated by the solid lines in (a) to (d) in fig. 5 are the actual water usage curves of the area on tuesday, thursday, saturday, and sunday, respectively; the curves of estimated water consumption for the area indicated by the dotted lines in the graphs (e) to (h) in fig. 5 are curves of estimated water consumption for the area on wednesday, thursday, saturday, sunday predicted from the water consumption of the previous day, respectively, that shows the predicted water consumption for the next 24 hours predicted from the actual water consumption on wednesday, friday, and saturday, respectively; the actual water usage curves indicated by the solid lines in (e) to (h) in fig. 5 are the actual water usage curves of the area on wednesday, thursday, saturday, and sunday, respectively.
As can be seen from the curves in the graph, the predicted data can basically keep the same variation trend as the actual data, but the farther the time for acquiring the water consumption in the previous time period is, the more the predicted data is deviated, for example, the difference between the estimated water consumption value and the actual water consumption value in the 80 th to 100 th periods in fig. 5(c) is larger; therefore, in the prediction, in order to improve the prediction accuracy, the collected water consumption in the previous time period is as close as possible to the prediction time period, for example, the difference between the estimated Saturday water consumption value predicted according to the Friday water consumption and the actual Saturday water consumption value in FIG. 5(g) is small. The leakage identification method based on the neural network, which is adopted by the embodiment, collects the water consumption of the last time period in a certain period, so that the water consumption of the last time period is as close to the prediction time period as possible, the accuracy of the predicted water consumption estimation value is improved, the leakage identification can be better carried out, and the accuracy of the leakage identification is further improved.
Example 3
Corresponding to the above method embodiments, the present embodiment further provides a leakage identification apparatus based on a neural network, and the leakage identification apparatus based on a neural network described below and the leakage identification method based on a neural network described above may be referred to correspondingly.
Referring to fig. 6, the apparatus includes the following modules: the system comprises a data acquisition module 101, a database 102, a neural network identification module 103, a data processing module 104 and an alarm module 105;
the output end of the data acquisition module 101 is connected with the database 102, the database is respectively connected with the first input ends of the neural network identification module 103 and the data processing module 104, the output end of the neural network identification module 103 is connected with the second input end of the data processing module 104, and the output end of the data processing module 104 is connected with the input end of the alarm module 105.
The data acquisition module 101 is used for acquiring flowmeter data, meteorological information, an annual date value and an intraday minute value; the database 102 is used for storing data acquired by the data acquisition module; the neural network identification module 103 is used for acquiring a water consumption estimation value of a region in the next time period according to the related data of the region in the previous time period input by the database 102; the data processing module 104 is used for calculating a comparison value according to the estimated value of the water consumption of the region in the next time period and the actual water consumption of the region, and judging whether to control the alarm module 105 to execute the alarm operation; the alarm module 105 is used for executing alarm operation according to the output signal of the data processing module 104.
The leakage recognition device based on the neural network further comprises a neural network training module 106; the output end of the database 102 is connected to the input end of the neural network training module 106, and the output end of the neural network training module 106 is connected to the second input end of the neural network identification module 103. The neural network training module 106 uses a moving window method to extract historical collected data in the database 102 to train the neural network, so that the time period of the training data keeps a fixed distance from the current time.
By applying the device provided by the embodiment of the invention, the flowmeter data, the meteorological information, the annual date value and the intraday minute value are obtained, then the water consumption estimated value of the region in the next time period is obtained through the neural network identification module, finally the leakage phenomenon is identified through the data processing module, and the alarm module is controlled to execute the alarm operation. In addition, the internal parameters of the neural network model are updated in time through the neural network training module, and the recognition accuracy of the neural network is improved.
Example 4
Corresponding to the above method embodiment, the present embodiment further provides a leakage identification device based on a neural network, and a leakage identification device based on a neural network described below and a leakage identification method based on a neural network described above may be referred to correspondingly.
Referring to fig. 7, the neural network based leakage identifying apparatus includes:
a memory D l for storing computer programs;
a processor D2 for implementing the steps of the neural network based leak recognition method of the above method embodiments when executing the computer program.
Specifically, referring to fig. 8, a schematic structural diagram of the leakage identification device based on the neural network provided in the present embodiment is provided, where the leakage identification device based on the neural network may generate a relatively large difference due to different configurations or performances, and may include one or more processors (CPUs) 322 (e.g., one or more processors) and a memory 332, and one or more storage media 330 (e.g., one or more mass storage devices) storing an application 342 or data 344. Memory 332 and storage media 330 may be, among other things, transient storage or persistent storage. The program stored on the storage medium 330 may include one or more modules (not shown), each of which may include a series of instructions operating on a data processing device. Still further, the central processor 322 may be configured to communicate with the storage medium 330, and execute a series of instruction operations in the storage medium 330 on the neural network-based leak recognition device 301.
The neural network based leak identification apparatus 301 may also include one or more power sources 326, one or more wired or wireless network interfaces 350, one or more input-output interfaces 358, and/or one or more operating systems 341. Such as Windows Server, Mac OS XTM, UnixTM, LinuxTM, FreeBSDTM, etc.
The steps in the neural network-based leak recognition method described above may be implemented by the structure of the neural network-based leak recognition apparatus.
Example 5
Corresponding to the above method embodiments, the present embodiment further provides a readable storage medium, and a readable storage medium described below and a neural network-based leakage identification method described above may be referred to correspondingly.
A readable storage medium, on which a computer program is stored, which, when being executed by a processor, carries out the steps of the neural network based leakage identification method of the above-mentioned method embodiments.
The readable storage medium may be a usb disk, a removable hard disk, a Read-Only Memory (ROM), a Random Access Memory (RAM), a magnetic disk or an optical disk, and various other readable storage media capable of storing program codes.
The foregoing is merely a detailed description of specific embodiments of the invention and is not intended to limit the invention. Various alterations, modifications and improvements will occur to those skilled in the art without departing from the spirit and scope of the invention.
Claims (10)
1. A leakage identification method based on a neural network is characterized by comprising the following steps:
s1, acquiring flow meter data, meteorological information, a day-of-year value and a minute-of-day value;
s2, calculating model input information through the data acquired in the step S1, wherein the model input information comprises water consumption of a time period in the region, a meteorological function, a time period function and a working day function;
s3, processing the model input information by using the neural network model to obtain the estimated value of water consumption in the next time period region;
s4, calculating a comparison value according to the estimated value of the water consumption of the region in the next time period and the actual water consumption of the region, and executing the step S5 if the comparison value is larger than a preset threshold value; when the comparison value is less than or equal to the preset threshold value, returning to the step S1;
and S5, executing alarm operation when the comparison value is larger than the preset threshold value, and then returning to the step S1.
2. The neural network-based leakage identification method according to claim 1, wherein the weather information obtained in step S1 includes temperature, humidity or precipitation information;
in the step S2, the meteorological function represents meteorological information, including temperature values or humidity values, and is represented in the form of an array; the time period function represents an annual date time interval and an intra-day time interval and is obtained by calculating an annual date value and an intra-day minute value respectively; the working day function represents working day or holiday information and is represented by the ratio of the time length of the date to the next nearest holiday to the end time length of the next nearest holiday.
3. The method of claim 1, wherein the neural network model is a BP neural network model.
4. The method for identifying leakage based on neural network as claimed in claim 1, wherein the training method of neural network model is: and (4) extracting historical acquisition data by adopting a moving window method to train the neural network.
5. The method for identifying leakage based on neural network as claimed in claim 1, wherein said step S5 specifically comprises the following steps:
s51, when the comparison value is larger than the preset threshold value, starting timing operation;
s52, in the timing time interval, if the time interval in which the comparison value exceeds the preset threshold value is larger than the preset time interval value, executing the step S53; otherwise, returning to the step S1;
and S53, executing alarm operation.
6. The neural network-based leak recognition method of claim 5, wherein the comparison value exceeds the predetermined threshold value in an inverse time-function relationship with the predetermined time interval value.
7. A leakage identification device based on a neural network is characterized by comprising a data acquisition module, a database, a neural network identification module, a data processing module and an alarm module;
the output end of the data acquisition module is connected with a database, the database is respectively connected with the first input ends of the neural network identification module and the data processing module, the output end of the neural network identification module is connected with the second input end of the data processing module, and the output end of the data processing module is connected with the input end of the alarm module;
the data acquisition module is used for acquiring flowmeter data, meteorological information, an annual date value and an intraday minute value; the database is used for storing the data acquired by the data acquisition module; the neural network identification module is used for acquiring a water consumption estimation value of a region in the next time period according to related data of the region in the previous time period input by the database; the data processing module is used for calculating a comparison value according to the estimated value of the water consumption of the region in the next time period and the actual water consumption of the region, and judging whether to control the alarm module to execute alarm operation or not; the alarm module is used for executing alarm operation according to the output signal of the data processing module.
8. The neural network-based leakage recognition device of claim 7, further comprising a neural network training module; the output end of the database is connected with the input end of the neural network training module, and the output end of the neural network training module is connected with the second input end of the neural network identification module; the neural network training module adopts a moving window method to extract historical acquisition data in a database to train the neural network.
9. A neural network-based leak recognition apparatus, comprising:
a memory for storing a computer program;
a processor for implementing the steps of the neural network based leak identification method of any one of claims 1 to 6 when executing the computer program.
10. A readable storage medium, having stored thereon a computer program which, when being executed by a processor, carries out the steps of the neural network based leakage identification method according to any one of claims 1 to 6.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911093678.8A CN110837933A (en) | 2019-11-11 | 2019-11-11 | Leakage identification method, device, equipment and storage medium based on neural network |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911093678.8A CN110837933A (en) | 2019-11-11 | 2019-11-11 | Leakage identification method, device, equipment and storage medium based on neural network |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN110837933A true CN110837933A (en) | 2020-02-25 |
Family
ID=69576350
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201911093678.8A Pending CN110837933A (en) | 2019-11-11 | 2019-11-11 | Leakage identification method, device, equipment and storage medium based on neural network |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110837933A (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113588179A (en) * | 2021-06-24 | 2021-11-02 | 武汉众智鸿图科技有限公司 | Water supply pipe network leakage detection method and system |
| CN113720425A (en) * | 2021-08-31 | 2021-11-30 | 福建蓝密码物联网科技有限公司 | Water leakage monitoring method and system based on intelligent water meter |
| CN113758627A (en) * | 2021-08-16 | 2021-12-07 | 杭州电子科技大学 | Transient flow event detection method for water supply pipe network |
| CN117726324A (en) * | 2024-02-07 | 2024-03-19 | 中国水利水电第九工程局有限公司 | Road traffic construction inspection method and system based on data identification |
Citations (13)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2000003353A (en) * | 1998-06-12 | 2000-01-07 | Toyo Electric Mfg Co Ltd | Water distribution prediction method |
| CN101008992A (en) * | 2006-12-30 | 2007-08-01 | 北京市劳动保护科学研究所 | Method for detecting leakage of pipeline based on artificial neural network |
| CN104044413A (en) * | 2013-03-14 | 2014-09-17 | 固特异轮胎和橡胶公司 | Predictive peer-based tire health monitoring |
| CN104061445A (en) * | 2014-07-09 | 2014-09-24 | 中国石油大学(华东) | Pipeline leakage detection method based on neural network |
| CN105678417A (en) * | 2016-01-06 | 2016-06-15 | 中电建路桥集团有限公司 | Prediction method and device for tunnel face water inflow of construction tunnel |
| CN106022518A (en) * | 2016-05-17 | 2016-10-12 | 清华大学 | Pipe damage probability prediction method based on BP neural network |
| CN108051035A (en) * | 2017-10-24 | 2018-05-18 | 清华大学 | The pipe network model recognition methods of neural network model based on gating cycle unit |
| CN109307159A (en) * | 2018-08-21 | 2019-02-05 | 湖南大学 | A kind of pipe network model alarm method based on water consumption optimal prediction model |
| CN109359698A (en) * | 2018-10-30 | 2019-02-19 | 清华大学 | Leakage loss recognition methods based on long Memory Neural Networks model in short-term |
| CN109555979A (en) * | 2018-12-10 | 2019-04-02 | 清华大学 | A kind of water supply network leakage monitoring method |
| CN109558900A (en) * | 2018-11-16 | 2019-04-02 | 佛山科学技术学院 | A kind of water supply pipe explosion time forecasting methods neural network based and device |
| CN109578816A (en) * | 2018-12-14 | 2019-04-05 | 中冶京诚工程技术有限公司 | Water supply network booster detection method and device, control method and device |
| MY169994A (en) * | 2012-10-02 | 2019-06-19 | Ranhill Water Services Sdn Bhd | A systems for monitoring and identifying nrw in a water utility network |
-
2019
- 2019-11-11 CN CN201911093678.8A patent/CN110837933A/en active Pending
Patent Citations (13)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2000003353A (en) * | 1998-06-12 | 2000-01-07 | Toyo Electric Mfg Co Ltd | Water distribution prediction method |
| CN101008992A (en) * | 2006-12-30 | 2007-08-01 | 北京市劳动保护科学研究所 | Method for detecting leakage of pipeline based on artificial neural network |
| MY169994A (en) * | 2012-10-02 | 2019-06-19 | Ranhill Water Services Sdn Bhd | A systems for monitoring and identifying nrw in a water utility network |
| CN104044413A (en) * | 2013-03-14 | 2014-09-17 | 固特异轮胎和橡胶公司 | Predictive peer-based tire health monitoring |
| CN104061445A (en) * | 2014-07-09 | 2014-09-24 | 中国石油大学(华东) | Pipeline leakage detection method based on neural network |
| CN105678417A (en) * | 2016-01-06 | 2016-06-15 | 中电建路桥集团有限公司 | Prediction method and device for tunnel face water inflow of construction tunnel |
| CN106022518A (en) * | 2016-05-17 | 2016-10-12 | 清华大学 | Pipe damage probability prediction method based on BP neural network |
| CN108051035A (en) * | 2017-10-24 | 2018-05-18 | 清华大学 | The pipe network model recognition methods of neural network model based on gating cycle unit |
| CN109307159A (en) * | 2018-08-21 | 2019-02-05 | 湖南大学 | A kind of pipe network model alarm method based on water consumption optimal prediction model |
| CN109359698A (en) * | 2018-10-30 | 2019-02-19 | 清华大学 | Leakage loss recognition methods based on long Memory Neural Networks model in short-term |
| CN109558900A (en) * | 2018-11-16 | 2019-04-02 | 佛山科学技术学院 | A kind of water supply pipe explosion time forecasting methods neural network based and device |
| CN109555979A (en) * | 2018-12-10 | 2019-04-02 | 清华大学 | A kind of water supply network leakage monitoring method |
| CN109578816A (en) * | 2018-12-14 | 2019-04-05 | 中冶京诚工程技术有限公司 | Water supply network booster detection method and device, control method and device |
Non-Patent Citations (2)
| Title |
|---|
| 付恒阳: "城市供水管网水漏损管理技术研究现状综述", 《陕西理工学院学报(自然科学版)》 * |
| 王婷婷等: "供水管网漏损管理研究现状与进展", 《供水技术》 * |
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113588179A (en) * | 2021-06-24 | 2021-11-02 | 武汉众智鸿图科技有限公司 | Water supply pipe network leakage detection method and system |
| CN113588179B (en) * | 2021-06-24 | 2023-11-21 | 武汉众智鸿图科技有限公司 | Water supply network leakage detection method and system |
| CN113758627A (en) * | 2021-08-16 | 2021-12-07 | 杭州电子科技大学 | Transient flow event detection method for water supply pipe network |
| CN113758627B (en) * | 2021-08-16 | 2024-01-02 | 杭州电子科技大学 | Water supply network transient flow event detection method |
| CN113720425A (en) * | 2021-08-31 | 2021-11-30 | 福建蓝密码物联网科技有限公司 | Water leakage monitoring method and system based on intelligent water meter |
| CN117726324A (en) * | 2024-02-07 | 2024-03-19 | 中国水利水电第九工程局有限公司 | Road traffic construction inspection method and system based on data identification |
| CN117726324B (en) * | 2024-02-07 | 2024-04-30 | 中国水利水电第九工程局有限公司 | Road traffic construction inspection method and system based on data identification |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110837933A (en) | Leakage identification method, device, equipment and storage medium based on neural network | |
| CN110376457B (en) | Non-invasive load monitoring method and device based on semi-supervised learning algorithm | |
| US11036189B2 (en) | Energy disaggregation techniques for low resolution whole-house energy consumption data | |
| Zhang et al. | An IoT-based thermal model learning framework for smart buildings | |
| CN109116444B (en) | PCA-kNN-based air quality model PM2.5Forecasting method | |
| CN109726865A (en) | Customer charge probability density prediction technique, device and storage medium based on EMD-QRF | |
| US20180039895A1 (en) | Data predicting method and apparatus | |
| CN102531121A (en) | Optimum input forecast system of water treatment coagulant and forecast method | |
| CN109598052B (en) | Intelligent ammeter life cycle prediction method and device based on correlation coefficient analysis | |
| US20210334914A1 (en) | System and method for determining power production in an electrical power grid | |
| US20230324860A1 (en) | Estimating Energy Consumption for a Building Using Dilated Convolutional Neural Networks | |
| CN111626006A (en) | Quantitative identification method for multivariate evaluation on runoff change sensitivity | |
| CN110751312A (en) | Multi-factor-based system dynamics life water demand prediction method and system | |
| CN112283593A (en) | Internet of things system for closing valve and detecting leakage of pipe network and leakage detection method thereof | |
| CN117878979A (en) | Power balance and dynamic compensation system based on electric energy storage | |
| JP2015137866A (en) | Water leakage amount estimation device, method, and system | |
| Boudhaouia et al. | Comparison of machine learning algorithms to predict daily water consumptions | |
| US20140203949A1 (en) | Method and system for automatic residual consumption | |
| CN113642248B (en) | Method and device for evaluating residual use time of positioning equipment | |
| CN111859783A (en) | Water pressure prediction method, system, storage medium, equipment and urban water supply system | |
| CN110322063A (en) | A kind of power consumption simulated prediction method and storage medium | |
| CN118094380B (en) | High-pressure oxygen therapy device use optimization method and system based on big data analysis | |
| CN113222266B (en) | Prediction model matching method based on large daily water meter data | |
| CN115719083A (en) | Model training method and device, computer equipment and readable storage medium | |
| CN118074126B (en) | Operation and maintenance monitoring method based on power grid information |
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 | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20200225 |