CN113551156A - Pipeline state monitoring method and device based on deep learning and storage medium - Google Patents

Pipeline state monitoring method and device based on deep learning and storage medium Download PDF

Info

Publication number
CN113551156A
CN113551156A CN202110698397.6A CN202110698397A CN113551156A CN 113551156 A CN113551156 A CN 113551156A CN 202110698397 A CN202110698397 A CN 202110698397A CN 113551156 A CN113551156 A CN 113551156A
Authority
CN
China
Prior art keywords
deep learning
water supply
state data
pipeline
learning model
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Granted
Application number
CN202110698397.6A
Other languages
Chinese (zh)
Other versions
CN113551156B (en
Inventor
林凡
黄富铿
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
GCI Science and Technology Co Ltd
Original Assignee
GCI Science and Technology Co Ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by GCI Science and Technology Co Ltd filed Critical GCI Science and Technology Co Ltd
Priority to CN202110698397.6A priority Critical patent/CN113551156B/en
Publication of CN113551156A publication Critical patent/CN113551156A/en
Application granted granted Critical
Publication of CN113551156B publication Critical patent/CN113551156B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • F—MECHANICAL ENGINEERING; LIGHTING; HEATING; WEAPONS; BLASTING
    • F17—STORING OR DISTRIBUTING GASES OR LIQUIDS
    • F17D—PIPE-LINE SYSTEMS; PIPE-LINES
    • F17D5/00—Protection or supervision of installations
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F17/00—Digital computing or data processing equipment or methods, specially adapted for specific functions
    • G06F17/10—Complex mathematical operations
    • G06F17/18—Complex mathematical operations for evaluating statistical data, e.g. average values, frequency distributions, probability functions, regression analysis
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/24—Classification techniques
    • G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
    • G06F18/2413—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on distances to training or reference patterns
    • G06F18/24133—Distances to prototypes
    • G06F18/24137—Distances to cluster centroïds
    • G06F18/2414—Smoothing the distance, e.g. radial basis function networks [RBFN]
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/24—Classification techniques
    • G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
    • G06F18/2415—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on parametric or probabilistic models, e.g. based on likelihood ratio or false acceptance rate versus a false rejection rate
    • 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
    • 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

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Data Mining & Analysis (AREA)
  • General Physics & Mathematics (AREA)
  • General Engineering & Computer Science (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Artificial Intelligence (AREA)
  • Evolutionary Computation (AREA)
  • Mathematical Physics (AREA)
  • Evolutionary Biology (AREA)
  • Software Systems (AREA)
  • Bioinformatics & Computational Biology (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • Biophysics (AREA)
  • Computational Mathematics (AREA)
  • Health & Medical Sciences (AREA)
  • Computational Linguistics (AREA)
  • General Health & Medical Sciences (AREA)
  • Molecular Biology (AREA)
  • Computing Systems (AREA)
  • Probability & Statistics with Applications (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Biomedical Technology (AREA)
  • Mathematical Analysis (AREA)
  • Mathematical Optimization (AREA)
  • Pure & Applied Mathematics (AREA)
  • Operations Research (AREA)
  • Algebra (AREA)
  • Databases & Information Systems (AREA)
  • Mechanical Engineering (AREA)
  • Pipeline Systems (AREA)
  • Examining Or Testing Airtightness (AREA)

Abstract

The invention discloses a pipeline state monitoring method, a device and a storage medium based on deep learning, wherein the method comprises the following steps: acquiring state data of each water supply pipeline according to a preset time interval to obtain state data sets of all the water supply pipelines, and dividing the state data sets into a training set and a testing set; wherein the status data comprises a pressure value, a flow value, and a pressure differential; inputting the training set into a deep learning model for training to obtain a trained deep learning model; acquiring real-time state data of each water supply pipeline in real time according to a preset time interval to obtain a real-time state data set; and inputting the real-time state data set into the trained deep learning model to obtain a pipeline state monitoring result. The embodiment of the invention is used for identifying and positioning the pipe burst of the water supply pipeline based on the existing low-frequency instrument equipment, has high detection precision and low false detection rate, and greatly improves the water supply safety guarantee capability.

Description

Pipeline state monitoring method and device based on deep learning and storage medium
Technical Field
The invention relates to the technical field of pipeline state monitoring, in particular to a pipeline state monitoring method and device based on deep learning and a storage medium.
Background
The analysis of related data shows that the leakage water quantity of a Chinese water business enterprise in one year is 102 hundred million tons calculated by the average leakage rate of 20 percent. If the cost of water supply is 1.5 yuan per ton of water, the direct economic loss due to leakage is 154 billion yuan per year. If the price of water sold per ton is 2 yuan, the economic loss of the leaked water amount is up to more than 200 million yuan. In addition, the leakage amount is further increased due to abnormal problems of overlarge water pressure, overlarge flow and the like, and the service life of a water supply pipeline is greatly shortened. Therefore, under the condition of not adopting large-area water cut-off, the abnormal water supply pipeline is positioned, thereby being beneficial to reducing leakage to the maximum extent and avoiding inducing serious safety accidents.
In the prior art, a low-frequency quasi-steady-state pressure and flow monitoring instrument and a negative pressure wave method are mostly adopted to monitor a water supply pipeline, however, the methods all need to adopt a high-precision high-frequency sensor, and are extremely inconvenient in practical engineering.
Disclosure of Invention
The technical problem to be solved by the embodiment of the invention is to provide a pipeline state monitoring method, a pipeline state monitoring device and a storage medium based on deep learning, wherein the pipeline explosion identification and positioning are carried out on a water supply pipeline based on the existing low-frequency instrument equipment, the detection precision is high, the false detection rate is low, and the water supply safety guarantee capability is greatly improved.
In order to achieve the above object, an embodiment of the present invention provides a method for monitoring a pipeline state based on deep learning, including:
acquiring state data of each water supply pipeline according to a preset time interval to obtain state data sets of all the water supply pipelines, and dividing the state data sets into a training set and a testing set; wherein the status data comprises a pressure value, a flow value, and a pressure differential;
inputting the training set into a deep learning model for training to obtain a trained deep learning model; the trained deep learning model comprises an input layer, a full connection layer and an output layer;
acquiring real-time state data of each water supply pipeline in real time according to a preset time interval to obtain a real-time state data set;
and inputting the real-time state data set into the trained deep learning model to obtain a pipeline state monitoring result.
As an improvement of the above scheme, the acquiring of the state data of each water supply pipeline according to a preset time interval obtains a state data set of all water supply pipelines, and divides the state data set into a training set and a testing set, specifically including:
collecting the pressure value of each pipeline according to a preset time interval
Figure BDA0003128731110000021
And flow rate value
Figure BDA0003128731110000022
Wherein,
Figure BDA0003128731110000023
denotes the nth (N ∈ [1, N)]) The value of the mth pressure of the strip line,
Figure BDA0003128731110000024
denotes the nth (N ∈ [1, N)]) The mth flow value of the strip pipe;
respectively calculating the average pressure value of each pipeline according to a preset time interval
Figure BDA0003128731110000025
Wherein,
Figure BDA0003128731110000026
an average pressure value representing the mth pressure value of the nth pipe;
calculating the pressure difference of each water supply pipeline according to the pressure value and the average pressure value
Figure BDA0003128731110000027
Wherein,
Figure BDA0003128731110000028
represents the mth pressure difference of the nth pipe;
acquiring the state data of each water supply pipeline according to a preset time interval
Figure BDA0003128731110000029
Wherein,
Figure BDA00031287311100000210
a state data set of all water supply pipes is obtained
Figure BDA00031287311100000211
Wherein,
Figure BDA00031287311100000212
the state data set is
Figure BDA00031287311100000213
And dividing the training set and the test set according to a preset proportion.
As an improvement of the above solution, the state data set
Figure BDA00031287311100000214
Before dividing into training set and test set according to the preset proportion, also include:
the state data set is
Figure BDA0003128731110000031
State data in
Figure BDA0003128731110000032
And performing label classification according to the normal data and the abnormal data.
As an improvement of the above scheme, the inputting the training set into a deep learning model for training to obtain a trained deep learning model specifically includes:
inputting the training set into a deep learning model for training;
and testing the trained deep learning model by using the test set to obtain the trained deep learning model.
As an improvement of the above scheme, the trained deep learning model comprises an input layer, three full-connection layers and an output layer; wherein,
the number of neurons of the input layer is 3N, the number of neurons is the same as that of water supply pipelines, and the activation function of the input layer is ReLu;
each fully connected layer comprises 1024 neurons, and the activation function of the fully connected layer is ReLu;
the number of neurons of the output layer is 3N +1, and the activation function of the output layer is Softmax.
The embodiment of the invention also provides a pipeline state monitoring device based on deep learning, which comprises:
the system comprises a first acquisition module, a second acquisition module and a control module, wherein the first acquisition module is used for acquiring state data of each water supply pipeline according to a preset time interval to obtain state data sets of all the water supply pipelines and dividing the state data sets into a training set and a testing set; wherein the status data comprises a pressure value, a flow value, and a pressure differential;
the training module is used for inputting the training set into a deep learning model for training to obtain a trained deep learning model; the trained deep learning model comprises an input layer, a full connection layer and an output layer;
the second acquisition module is used for acquiring real-time state data of each water supply pipeline in real time according to a preset time interval to obtain a real-time state data set;
and the monitoring module is used for inputting the real-time state data set into the trained deep learning model to obtain a pipeline state monitoring result.
Further, the first obtaining module specifically includes:
a collecting unit for collecting pressure value of each pipeline according to preset time interval
Figure BDA0003128731110000041
And flow rate value
Figure BDA0003128731110000042
Wherein,
Figure BDA0003128731110000043
denotes the nth (N ∈ [1, N)]) The value of the mth pressure of the strip line,
Figure BDA0003128731110000044
denotes the nth (N ∈ [1, N)]) The mth flow value of the strip pipe;
a first calculating unit for calculating the average pressure value of each pipeline according to a preset time interval
Figure BDA0003128731110000045
Wherein,
Figure BDA0003128731110000046
an average pressure value representing the mth pressure value of the nth pipe;
a second calculation unit for calculating the pressure difference of each water supply pipeline according to the pressure value and the average pressure value
Figure BDA0003128731110000047
Wherein,
Figure BDA0003128731110000048
represents the mth pressure difference of the nth pipe;
a data set acquisition unit for acquiring the state data of each water supply pipeline according to a preset time interval
Figure BDA0003128731110000049
Wherein,
Figure BDA00031287311100000410
a state data set of all water supply pipes is obtained
Figure BDA00031287311100000411
Wherein,
Figure BDA00031287311100000412
a classification unit for classifying the state data set
Figure BDA00031287311100000413
And dividing the training set and the test set according to a preset proportion.
Further, the trained deep learning model comprises an input layer, three full-connection layers and an output layer; wherein,
the number of neurons of the input layer is 3N, the number of neurons is the same as that of water supply pipelines, and the activation function of the input layer is ReLu;
each fully connected layer comprises 1024 neurons, and the activation function of the fully connected layer is ReLu;
the number of neurons of the output layer is 3N +1, and the activation function of the output layer is Softmax.
The embodiment of the present invention further provides a terminal device, which includes a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, and when the processor executes the computer program, the processor implements any one of the deep learning-based pipe state monitoring methods described above.
The embodiment of the invention also provides a computer-readable storage medium, which includes a stored computer program, wherein when the computer program runs, the apparatus where the computer-readable storage medium is located is controlled to execute any one of the above deep learning-based pipeline state monitoring methods.
Compared with the prior art, the pipeline state monitoring method, the pipeline state monitoring device and the storage medium based on deep learning provided by the embodiment of the invention have the beneficial effects that: acquiring state data of each water supply pipeline according to a preset time interval to obtain state data sets of all the water supply pipelines, and dividing the state data sets into a training set and a testing set; wherein the status data comprises a pressure value, a flow value, and a pressure differential; inputting the training set into a deep learning model for training to obtain a trained deep learning model; the trained deep learning model comprises an input layer, a full connection layer and an output layer; acquiring real-time state data of each water supply pipeline in real time according to a preset time interval to obtain a real-time state data set; and inputting the real-time state data set into the trained deep learning model to obtain a pipeline state monitoring result. The embodiment of the invention is used for identifying and positioning the pipe burst of the water supply pipeline based on the existing low-frequency instrument equipment, has high detection precision and low false detection rate, and greatly improves the water supply safety guarantee capability.
Drawings
FIG. 1 is a schematic flow chart diagram of a method for monitoring a pipeline state based on deep learning according to a preferred embodiment of the present invention;
FIG. 2 is an architecture diagram of a deep learning model in a preferred embodiment of a deep learning-based pipeline condition monitoring method provided by the invention;
FIG. 3 is a schematic structural diagram of a preferred embodiment of a deep learning-based pipeline condition monitoring device provided by the invention;
fig. 4 is a schematic structural diagram of a preferred embodiment of a terminal device provided by the present invention.
Detailed Description
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
Referring to fig. 1, fig. 1 is a schematic flow chart illustrating a method for monitoring a pipeline state based on deep learning according to a preferred embodiment of the present invention. The pipeline state monitoring method based on deep learning comprises the following steps:
s1, acquiring state data of each water supply pipeline according to a preset time interval to obtain state data sets of all the water supply pipelines, and dividing the state data sets into a training set and a testing set; wherein the status data comprises a pressure value, a flow value, and a pressure differential;
s2, inputting the training set into a deep learning model for training to obtain a trained deep learning model; the trained deep learning model comprises an input layer, a full connection layer and an output layer;
s3, acquiring real-time state data of each water supply pipeline in real time according to a preset time interval to obtain a real-time state data set;
and S4, inputting the real-time state data set into the trained deep learning model to obtain a pipeline state monitoring result.
Specifically, before state data of each water supply pipeline is acquired according to a preset time interval, the number of all water supply pipelines in a certain city is calculated; wherein, a water supply pipeline is calculated between every two nodes, and each water supply pipeline represents L if a total of N water supply pipelines is provided1、L2、......、LN(ii) a And a water pressure sensor and a flow sensor are arranged in each water supply pipeline for collecting the pressure value and the flow value of the water supply pipeline. Then acquiring state data of each water supply pipeline according to a preset time interval to obtain state data sets of all the water supply pipelines, and dividing the state data sets into a training set and a testing set; wherein the status data comprises a pressure value, a flow value, and a pressure differential; inputting the training set into a deep learning model for training to obtain a trained deep learning model; the trained deep learning model comprises an input layer, a full connection layer and an output layer; acquiring real-time state data of each water supply pipeline in real time according to a preset time interval to obtain a real-time state data set; and inputting the real-time state data set into the trained deep learning model to obtain a pipeline state monitoring result.
The embodiment identifies and positions pipe explosion of the water supply pipeline based on the existing low-frequency instrument equipment, has high detection precision and low false detection rate, and greatly improves the water supply safety guarantee capability; and meanwhile, the method also has better robustness under the condition that the data quantity is sufficient.
In another preferred embodiment, the S1, acquiring the status data of each water supply pipeline according to a preset time interval, obtaining a status data set of all water supply pipelines, and dividing the status data set into a training set and a testing set, specifically includes:
s101, collecting pressure values of all pipelines according to a preset time interval
Figure BDA0003128731110000071
And flow rate value
Figure BDA0003128731110000072
Wherein,
Figure BDA0003128731110000073
denotes the nth (N ∈ [1, N)]) The value of the mth pressure of the strip line,
Figure BDA0003128731110000074
denotes the nth (N ∈ [1, N)]) The mth flow value of the strip pipe;
s102, respectively calculating the average pressure value of each pipeline according to a preset time interval
Figure BDA0003128731110000075
Wherein,
Figure BDA0003128731110000076
an average pressure value representing the mth pressure value of the nth pipe;
s103, calculating the pressure difference of each water supply pipeline according to the pressure value and the average pressure value
Figure BDA0003128731110000077
Wherein,
Figure BDA0003128731110000078
indicates the nth tubeThe mth pressure difference of the lane;
s104, acquiring the state data of each water supply pipeline according to a preset time interval
Figure BDA0003128731110000079
Wherein,
Figure BDA00031287311100000710
a state data set of all water supply pipes is obtained
Figure BDA00031287311100000711
Wherein,
Figure BDA00031287311100000712
s105, collecting the state data set
Figure BDA00031287311100000713
And dividing the training set and the test set according to a preset proportion.
Specifically, the pressure value of each water supply pipeline is acquired by a sensor of the pipeline every 1 hour
Figure BDA00031287311100000714
And flow rate value
Figure BDA00031287311100000715
Wherein
Figure BDA00031287311100000716
Denotes the nth (N ∈ [1, N)]) The value of the mth pressure of the strip line,
Figure BDA00031287311100000717
denotes the nth (N ∈ [1, N)]) The mth flow value of the strip pipe; respectively calculating the average pressure value of each water supply pipeline for nearly 24 hours every 1 hour
Figure BDA00031287311100000718
Wherein,
Figure BDA00031287311100000719
an average pressure value representing the mth pressure value of the nth pipe; according to pressure value every 1 hour
Figure BDA00031287311100000720
Average pressure value
Figure BDA00031287311100000721
The pressure difference of each water supply pipeline is calculated respectively
Figure BDA00031287311100000722
Figure BDA00031287311100000723
Wherein,
Figure BDA00031287311100000724
the mth pressure difference of the nth pipe; acquiring the state data of each pipeline every 1 hour
Figure BDA00031287311100000725
Wherein
Figure BDA00031287311100000726
A state data set of all water supply lines is available every hour
Figure BDA00031287311100000727
Wherein,
Figure BDA00031287311100000728
the state data set is
Figure BDA00031287311100000729
And dividing the training set and the test set according to a preset proportion.
In yet another preferred embodiment, the S105, the state data set
Figure BDA0003128731110000081
Dividing the training set into a training set and a test according to a preset proportionBefore the collection, the method further comprises the following steps:
the state data set is
Figure BDA0003128731110000082
State data in
Figure BDA0003128731110000083
And performing label classification according to the normal data and the abnormal data.
Specifically, for example, 10000 sets of data are collected
Figure BDA0003128731110000084
And performing label classification on the normal data and the abnormal data, wherein the normal data is 2000 groups, the abnormal data is 8000 groups, and the normal data and the abnormal data must include data of all N pipelines. And then, disordering the data classified by all the labels, wherein 80% of the data is used as a training set, and 20% of the data is used as a testing set.
In another preferred embodiment, the S2, inputting the training set into a deep learning model for training, to obtain a trained deep learning model, specifically including:
s201, inputting the training set into a deep learning model for training;
and S202, testing the trained deep learning model by using the test set to obtain the trained deep learning model.
Specifically, the training set is input into a deep learning model for training, the Mean Absolute Error (MAE) is adopted to calculate the error (loss) in the training process, and an optimizer (optimizer) is preferably set as the adaptive moment estimation (Adam). Dropout is preferably set to 0.5, batch _ size is preferably set to 256, and epochs is preferably set to 1000. And after 1000 epochs, finishing the training of the deep learning model. And testing the trained deep learning model by using a test set to obtain the trained deep learning model. epochs are defined as a single training iteration of all batches in forward and backward propagation, meaning that 1 cycle is a single forward and backward pass of the entire input data. In brief, epochs refer to how many times the data will be "turned" during the training process. It should be noted that the error calculation method, the selection of the optimizer, and the setting of the parameters are not limited to those provided in the embodiment, and may be determined according to actual situations.
As a preferred scheme, the trained deep learning model comprises an input layer, three full-connection layers and an output layer; wherein,
the number of neurons of the input layer is 3N, the number of neurons is the same as that of water supply pipelines, and the activation function of the input layer is ReLu;
each fully connected layer comprises 1024 neurons, and the activation function of the fully connected layer is ReLu;
the number of neurons of the output layer is 3N +1, and the activation function of the output layer is Softmax.
Specifically, please refer to fig. 2, fig. 2 is an architecture diagram of a deep learning model in a preferred embodiment of a pipeline condition monitoring method based on deep learning according to the present invention. The trained deep learning model comprises an input layer, three full-connection layers and an output layer; wherein, the number of neurons in the input layer is 3N, and the corresponding vector Z &mThe number of the elements is the same as that of the water supply pipeline, and the activation function of the input layer is ReLu; each fully connected layer comprises 1024 neurons, and the activation function of the fully connected layer is ReLu; the number of neurons in the output layer is 3N +1, and the activation function of the output layer is Softmax, which means that a certain maximum value is not uniquely determined any more, but a probability value is assigned to each output classification result, indicating the possibility of belonging to each class. Collecting real-time state data of each water supply pipeline in real time every other hour to obtain a real-time state data set of all water supply pipelines
Figure BDA0003128731110000091
Collecting real-time status data
Figure BDA0003128731110000092
Inputting the trained deep learning model, the output layer will output different classification probabilities and arrangementGet up to P1P2…PN+1As shown in table 1 below, the category with the highest probability is selected as the output, and the pipeline state monitoring result is obtained. For example: if the pipeline 1 is broken, the flow rate and the pressure are abnormal, and the acquired data set
Figure BDA0003128731110000093
Inputting the data into a trained deep learning model, and outputting P by deep learning2Further, it is possible to determine that the pipeline 1 is abnormal, and to quickly locate the abnormal pipeline. At the moment, workers can be dispatched to check and further detect the pressure and flow sensors on the spot, if the condition of leakage and pipe explosion occurs, the pipeline is timely maintained, and if the pressure and flow sensors are checked without problems, the pressure and flow sensors can be replaced.
Inputting data Deep learning model output
Normal data P1
Pipeline 1 anomaly data P2
Pipe 2 anomaly data P3
...... ......
Pipeline N anomaly data PN+1
Table 1 deep learning output table
Correspondingly, the invention also provides a deep learning-based pipeline state monitoring device, which can realize all the processes of the deep learning-based pipeline state monitoring method in the embodiment.
Referring to fig. 3, fig. 3 is a schematic structural diagram of a pipeline condition monitoring device based on deep learning according to a preferred embodiment of the present invention. Pipeline state monitoring devices based on degree of depth study includes:
the first obtaining module 301 is configured to obtain state data of each water supply pipeline according to a preset time interval, obtain state data sets of all water supply pipelines, and divide the state data sets into a training set and a test set; wherein the status data comprises a pressure value, a flow value, and a pressure differential;
a training module 302, configured to input the training set into a deep learning model for training, so as to obtain a trained deep learning model; the trained deep learning model comprises an input layer, a full connection layer and an output layer;
a second obtaining module 303, configured to obtain real-time status data of each water supply pipeline in real time according to a preset time interval, so as to obtain a real-time status data set;
and the monitoring module 304 is configured to input the real-time state data set to the trained deep learning model to obtain a pipeline state monitoring result.
Preferably, the first obtaining module 301 specifically includes:
a collecting unit 311 for collecting the pressure value of each pipeline according to a preset time interval
Figure BDA0003128731110000108
And flow rate value
Figure BDA0003128731110000105
Wherein,
Figure BDA0003128731110000106
denotes the nth (N ∈ [1, N)]) The value of the mth pressure of the strip line,
Figure BDA0003128731110000107
denotes the nth (N ∈ [1, N)]) The mth flow value of the strip pipe;
a first calculating unit 312 for calculating an average pressure value of each pipeline according to a preset time interval
Figure BDA0003128731110000101
Wherein,
Figure BDA0003128731110000102
an average pressure value representing the mth pressure value of the nth pipe;
a second calculation unit 313 for calculating a pressure difference of each water supply pipeline according to the pressure value and the average pressure value
Figure BDA0003128731110000103
Wherein,
Figure BDA0003128731110000104
represents the mth pressure difference of the nth pipe;
a data set acquisition unit 314 for acquiring status data of each water supply pipeline at predetermined time intervals
Figure BDA0003128731110000111
Wherein,
Figure BDA0003128731110000112
a state data set of all water supply pipes is obtained
Figure BDA0003128731110000113
Wherein,
Figure BDA0003128731110000114
a classification unit 315 for classifyingSaid state data set
Figure BDA0003128731110000115
And dividing the training set and the test set according to a preset proportion.
Preferably, said state data set is
Figure BDA0003128731110000116
Before dividing into training set and test set according to the preset proportion, also include:
the state data set is
Figure BDA0003128731110000117
State data in
Figure BDA0003128731110000118
And performing label classification according to the normal data and the abnormal data.
Preferably, the training module 302 is specifically configured to:
inputting the training set into a deep learning model for training;
and testing the trained deep learning model by using the test set to obtain the trained deep learning model.
Preferably, the trained deep learning model comprises an input layer, three full-connection layers and an output layer; wherein,
the number of neurons of the input layer is 3N, the number of neurons is the same as that of water supply pipelines, and the activation function of the input layer is ReLu;
each fully connected layer comprises 1024 neurons, and the activation function of the fully connected layer is ReLu;
the number of neurons of the output layer is 3N +1, and the activation function of the output layer is Softmax.
In a specific implementation, the working principle, the control flow and the technical effect of the deep learning-based pipeline state monitoring device provided in the embodiment of the present invention are the same as those of the deep learning-based pipeline state monitoring method in the above embodiment, and are not described herein again.
Referring to fig. 4, fig. 4 is a schematic structural diagram of a terminal device according to a preferred embodiment of the present invention. The terminal device comprises a processor 401, a memory 402 and a computer program stored in the memory 402 and configured to be executed by the processor 401, wherein the processor 401 implements the deep learning based pipe state monitoring method according to any one of the above embodiments when executing the computer program.
Preferably, the computer program may be divided into one or more modules/units (e.g., computer program 1, computer program 2, … …) that are stored in the memory 402 and executed by the processor 401 to implement the invention. The one or more modules/units may be a series of computer program instruction segments capable of performing specific functions, which are used for describing the execution process of the computer program in the terminal device.
The Processor 401 may be a Central Processing Unit (CPU), other general purpose Processor, a Digital Signal Processor (DSP), an Application Specific Integrated Circuit (ASIC), a Field Programmable Gate Array (FPGA) or other Programmable logic device, a discrete Gate or transistor logic device, a discrete hardware component, etc., the general purpose Processor may be a microprocessor, or the Processor 401 may be any conventional Processor, the Processor 401 is a control center of the terminal device, and various interfaces and lines are used to connect various parts of the terminal device.
The memory 402 mainly includes a program storage area and a data storage area, wherein the program storage area may store an operating system, an application program required for at least one function, and the like, and the data storage area may store related data and the like. In addition, the memory 402 may be a high speed random access memory, a non-volatile memory such as a plug-in hard disk, a Smart Media Card (SMC), a Secure Digital (SD) Card, a Flash Card (Flash Card), and the like, or the memory 402 may be other volatile solid state memory devices.
It should be noted that the terminal device may include, but is not limited to, a processor and a memory, and those skilled in the art will understand that the structural diagram of fig. 4 is only an example of the terminal device and does not constitute a limitation of the terminal device, and may include more or less components than those shown, or combine some components, or different components.
The embodiment of the present invention further provides a computer-readable storage medium, where the computer-readable storage medium includes a stored computer program, and when the computer program runs, a device where the computer-readable storage medium is located is controlled to execute the deep learning-based pipeline state monitoring method according to any one of the above embodiments.
The embodiment of the invention provides a pipeline state monitoring method, a pipeline state monitoring device and a storage medium based on deep learning, wherein a state data set of all water supply pipelines is obtained by acquiring state data of each water supply pipeline according to a preset time interval, and the state data set is divided into a training set and a testing set; wherein the status data comprises a pressure value, a flow value, and a pressure differential; inputting the training set into a deep learning model for training to obtain a trained deep learning model; the trained deep learning model comprises an input layer, a full connection layer and an output layer; acquiring real-time state data of each water supply pipeline in real time according to a preset time interval to obtain a real-time state data set; and inputting the real-time state data set into the trained deep learning model to obtain a pipeline state monitoring result. The embodiment of the invention is used for identifying and positioning the pipe burst of the water supply pipeline based on the existing low-frequency instrument equipment, has high detection precision and low false detection rate, and greatly improves the water supply safety guarantee capability.
It should be noted that the above-described system embodiments are merely illustrative, where the units described as separate parts may or may not be physically separate, and the parts displayed as units may or may not be physical units, may be located in one place, or may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of the present embodiment. In addition, in the drawings of the embodiment of the system provided by the present invention, the connection relationship between the modules indicates that there is a communication connection between them, and may be specifically implemented as one or more communication buses or signal lines. One of ordinary skill in the art can understand and implement it without inventive effort.
While the foregoing is directed to the preferred embodiment of the present invention, it will be understood by those skilled in the art that various changes and modifications may be made without departing from the spirit and scope of the invention.

Claims (10)

1. A pipeline state monitoring method based on deep learning is characterized by comprising the following steps:
acquiring state data of each water supply pipeline according to a preset time interval to obtain state data sets of all the water supply pipelines, and dividing the state data sets into a training set and a testing set; wherein the status data comprises a pressure value, a flow value, and a pressure differential;
inputting the training set into a deep learning model for training to obtain a trained deep learning model; the trained deep learning model comprises an input layer, a full connection layer and an output layer;
acquiring real-time state data of each water supply pipeline in real time according to a preset time interval to obtain a real-time state data set;
and inputting the real-time state data set into the trained deep learning model to obtain a pipeline state monitoring result.
2. The pipeline condition monitoring method based on deep learning of claim 1, wherein the acquiring of the condition data of each water supply pipeline at preset time intervals obtains a condition data set of all water supply pipelines, and the dividing of the condition data set into a training set and a testing set specifically comprises:
collecting the pressure value of each pipeline according to a preset time interval
Figure FDA0003128731100000011
And flow rate value
Figure FDA0003128731100000012
Wherein,
Figure FDA0003128731100000013
denotes the nth (N ∈ [1, N)]) The value of the mth pressure of the strip line,
Figure FDA0003128731100000014
denotes the nth (N ∈ [1, N)]) The mth flow value of the strip pipe;
respectively calculating the average pressure value of each pipeline according to a preset time interval
Figure FDA0003128731100000015
Wherein,
Figure FDA0003128731100000016
an average pressure value representing the mth pressure value of the nth pipe;
calculating the pressure difference of each water supply pipeline according to the pressure value and the average pressure value
Figure FDA0003128731100000017
Wherein,
Figure FDA0003128731100000018
represents the mth pressure difference of the nth pipe;
acquiring the state data of each water supply pipeline according to a preset time interval
Figure FDA0003128731100000019
Wherein,
Figure FDA0003128731100000021
a state data set of all water supply pipes is obtained
Figure FDA0003128731100000022
Wherein,
Figure FDA0003128731100000023
the state data set is
Figure FDA0003128731100000024
And dividing the training set and the test set according to a preset proportion.
3. The deep learning-based pipeline condition monitoring method according to claim 2, wherein the condition data set is
Figure FDA0003128731100000025
Before dividing into training set and test set according to the preset proportion, also include:
the state data set is
Figure FDA0003128731100000026
State data in
Figure FDA0003128731100000027
And performing label classification according to the normal data and the abnormal data.
4. The method for monitoring the state of the pipeline based on the deep learning of claim 1, wherein the step of inputting the training set into a deep learning model for training to obtain the trained deep learning model specifically comprises the steps of:
inputting the training set into a deep learning model for training;
and testing the trained deep learning model by using the test set to obtain the trained deep learning model.
5. The deep learning-based pipeline condition monitoring method according to claim 4, wherein the trained deep learning model comprises an input layer, three fully-connected layers and an output layer; wherein,
the number of neurons of the input layer is 3N, the number of neurons is the same as that of water supply pipelines, and the activation function of the input layer is ReLu;
each fully connected layer comprises 1024 neurons, and the activation function of the fully connected layer is ReLu;
the number of neurons of the output layer is 3N +1, and the activation function of the output layer is Softmax.
6. A pipe condition monitoring device based on deep learning, comprising:
the system comprises a first acquisition module, a second acquisition module and a control module, wherein the first acquisition module is used for acquiring state data of each water supply pipeline according to a preset time interval to obtain state data sets of all the water supply pipelines and dividing the state data sets into a training set and a testing set; wherein the status data comprises a pressure value, a flow value, and a pressure differential;
the training module is used for inputting the training set into a deep learning model for training to obtain a trained deep learning model; the trained deep learning model comprises an input layer, a full connection layer and an output layer;
the second acquisition module is used for acquiring real-time state data of each water supply pipeline in real time according to a preset time interval to obtain a real-time state data set;
and the monitoring module is used for inputting the real-time state data set into the trained deep learning model to obtain a pipeline state monitoring result.
7. The deep learning-based pipeline condition monitoring device according to claim 6, wherein the first obtaining module specifically comprises:
a collecting unit for collecting each tube at a predetermined time intervalPressure value of road
Figure FDA0003128731100000031
And flow rate value
Figure FDA0003128731100000032
Wherein,
Figure FDA0003128731100000033
denotes the nth (N ∈ [1, N)]) The value of the mth pressure of the strip line,
Figure FDA0003128731100000034
denotes the nth (N ∈ [1, N)]) The mth flow value of the strip pipe;
a first calculating unit for calculating the average pressure value of each pipeline according to a preset time interval
Figure FDA0003128731100000035
Wherein,
Figure FDA0003128731100000036
an average pressure value representing the mth pressure value of the nth pipe;
a second calculation unit for calculating the pressure difference of each water supply pipeline according to the pressure value and the average pressure value
Figure FDA0003128731100000037
Wherein,
Figure FDA0003128731100000038
represents the mth pressure difference of the nth pipe;
a data set acquisition unit for acquiring the state data of each water supply pipeline according to a preset time interval
Figure FDA0003128731100000039
Wherein,
Figure FDA00031287311000000310
a state data set of all water supply pipes is obtained
Figure FDA00031287311000000311
Wherein,
Figure FDA0003128731100000041
a classification unit for classifying the state data set
Figure FDA0003128731100000042
And dividing the training set and the test set according to a preset proportion.
8. The deep learning-based pipeline condition monitoring device of claim 7, wherein the trained deep learning model comprises an input layer, three fully-connected layers and an output layer; wherein,
the number of neurons of the input layer is 3N, the number of neurons is the same as that of water supply pipelines, and the activation function of the input layer is ReLu;
each fully connected layer comprises 1024 neurons, and the activation function of the fully connected layer is ReLu;
the number of neurons of the output layer is 3N +1, and the activation function of the output layer is Softmax.
9. A terminal device comprising a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, the processor implementing the deep learning based pipe condition monitoring method according to any one of claims 1 to 5 when executing the computer program.
10. A computer-readable storage medium, comprising a stored computer program, wherein the computer program, when executed, controls an apparatus in which the computer-readable storage medium is located to perform the deep learning-based pipeline condition monitoring method according to any one of claims 1 to 5.
CN202110698397.6A 2021-06-23 2021-06-23 Pipeline state monitoring method and device based on deep learning and storage medium Active CN113551156B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202110698397.6A CN113551156B (en) 2021-06-23 2021-06-23 Pipeline state monitoring method and device based on deep learning and storage medium

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202110698397.6A CN113551156B (en) 2021-06-23 2021-06-23 Pipeline state monitoring method and device based on deep learning and storage medium

Publications (2)

Publication Number Publication Date
CN113551156A true CN113551156A (en) 2021-10-26
CN113551156B CN113551156B (en) 2023-04-07

Family

ID=78130820

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202110698397.6A Active CN113551156B (en) 2021-06-23 2021-06-23 Pipeline state monitoring method and device based on deep learning and storage medium

Country Status (1)

Country Link
CN (1) CN113551156B (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2024069216A1 (en) * 2022-09-30 2024-04-04 Matrix Jvco Ltd Trading As Aiq Method for pipeline corrosion prediction and maintenance

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20180365555A1 (en) * 2016-12-22 2018-12-20 Naveed Aslam Artificial intelligence based algorithm for predicting pipeline leak and corrosion detection
CN109242049A (en) * 2018-11-21 2019-01-18 安徽建筑大学 Water supply pipe network multipoint leakage positioning method and device based on convolutional neural network
CN110145695A (en) * 2019-06-03 2019-08-20 大连理工大学 A kind of hot duct leakage detection method based on the fusion of the depth confidence network information
CN111425768A (en) * 2020-03-31 2020-07-17 长安大学 Oil pipeline oil leakage point and oil leakage rate detection method based on underwater sensor network
CN112610903A (en) * 2020-12-10 2021-04-06 合肥学院 Water supply pipe network leakage positioning method based on deep neural network model
CN112949817A (en) * 2020-12-24 2021-06-11 长江勘测规划设计研究有限责任公司 Water supply pipe leakage edge equipment detection method based on time convolution neural network

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20180365555A1 (en) * 2016-12-22 2018-12-20 Naveed Aslam Artificial intelligence based algorithm for predicting pipeline leak and corrosion detection
CN109242049A (en) * 2018-11-21 2019-01-18 安徽建筑大学 Water supply pipe network multipoint leakage positioning method and device based on convolutional neural network
CN110145695A (en) * 2019-06-03 2019-08-20 大连理工大学 A kind of hot duct leakage detection method based on the fusion of the depth confidence network information
CN111425768A (en) * 2020-03-31 2020-07-17 长安大学 Oil pipeline oil leakage point and oil leakage rate detection method based on underwater sensor network
CN112610903A (en) * 2020-12-10 2021-04-06 合肥学院 Water supply pipe network leakage positioning method based on deep neural network model
CN112949817A (en) * 2020-12-24 2021-06-11 长江勘测规划设计研究有限责任公司 Water supply pipe leakage edge equipment detection method based on time convolution neural network

Non-Patent Citations (2)

* Cited by examiner, † Cited by third party
Title
牟天蔚等: "基于深度学习框架的供水管网压力异常实时诊断方法", 《水电能源科学》 *
王新颖等: "深度学习神经网络在管道故障诊断中的应用研究", 《安全与环境工程》 *

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2024069216A1 (en) * 2022-09-30 2024-04-04 Matrix Jvco Ltd Trading As Aiq Method for pipeline corrosion prediction and maintenance

Also Published As

Publication number Publication date
CN113551156B (en) 2023-04-07

Similar Documents

Publication Publication Date Title
CN108593260B (en) Optical cable line fault positioning and detecting method and terminal equipment
CN106897178B (en) Slow disk detection method and system based on extreme learning machine
CN111459700A (en) Method and apparatus for diagnosing device failure, diagnostic device, and storage medium
CN111078552A (en) Method and device for detecting page display abnormity and storage medium
CN113592019A (en) Fault detection method, device, equipment and medium based on multi-model fusion
CN104598342A (en) Internal storage detection method and device
CN112097125B (en) Water supply pipe network pipe burst detection and positioning method based on self-adaptive checking
CN113551156B (en) Pipeline state monitoring method and device based on deep learning and storage medium
CN113282000A (en) Fault diagnosis method and device of data center and dynamic loop monitoring system
CN116593897A (en) Power battery fault diagnosis method, system, vehicle and storage medium
CN117170304B (en) PLC remote monitoring control method and system based on industrial Internet of things
CN117114248B (en) Health evaluation method for water quality monitoring system, computer equipment and medium
CN114417995A (en) Pipe network risk early warning method and device, computer equipment and storage medium
CN112949697A (en) Method and device for confirming pipeline abnormity and computer readable storage medium
CN116484219A (en) Water supply network water quality abnormal pollution source identification method based on gate control graph neural network
CN115330088A (en) Flood control fine prediction method for small reservoir
CN117454563A (en) Leakage detection coverage rate detection method and device based on risk frequency
CN114416467A (en) Anomaly detection method and device
CN113779926A (en) Circuit detection method and device, electronic equipment and readable storage medium
CN112906727A (en) Method and system for real-time online detection of virtual machine state
CN111176931A (en) Operation monitoring method, operation monitoring device, server and storage medium
CN111026608A (en) Oracle database performance diagnosis method and system, equipment and medium thereof
CN118094450B (en) Fault early warning method and related equipment
US20120166888A1 (en) Method and software program product for determining the quality of an information technology system
CN113515507B (en) Method and system applied to dam water seepage detection

Legal Events

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