CN115310258A - Underground pipeline service life prediction method based on LSTM - Google Patents
Underground pipeline service life prediction method based on LSTM Download PDFInfo
- Publication number
- CN115310258A CN115310258A CN202210729622.2A CN202210729622A CN115310258A CN 115310258 A CN115310258 A CN 115310258A CN 202210729622 A CN202210729622 A CN 202210729622A CN 115310258 A CN115310258 A CN 115310258A
- Authority
- CN
- China
- Prior art keywords
- lstm
- data
- training
- model
- prediction method
- 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
- 238000000034 method Methods 0.000 title claims abstract description 30
- 238000012549 training Methods 0.000 claims abstract description 45
- 238000012360 testing method Methods 0.000 claims abstract description 29
- 238000012544 monitoring process Methods 0.000 claims abstract description 17
- 238000004422 calculation algorithm Methods 0.000 claims abstract description 7
- 238000007781 pre-processing Methods 0.000 claims abstract description 7
- ORILYTVJVMAKLC-UHFFFAOYSA-N Adamantane Natural products C1C(C2)CC3CC1CC2C3 ORILYTVJVMAKLC-UHFFFAOYSA-N 0.000 claims abstract description 6
- 230000009471 action Effects 0.000 claims abstract description 6
- 238000011156 evaluation Methods 0.000 claims abstract description 6
- 238000005457 optimization Methods 0.000 claims abstract description 6
- 230000009466 transformation Effects 0.000 claims abstract description 6
- 230000002159 abnormal effect Effects 0.000 claims description 6
- 125000004122 cyclic group Chemical group 0.000 claims description 3
- 210000002569 neuron Anatomy 0.000 claims description 3
- 238000012545 processing Methods 0.000 claims description 3
- 238000004364 calculation method Methods 0.000 claims description 2
- 238000013528 artificial neural network Methods 0.000 abstract description 3
- 230000008569 process Effects 0.000 description 6
- 230000007797 corrosion Effects 0.000 description 2
- 238000005260 corrosion Methods 0.000 description 2
- 230000007547 defect Effects 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 230000006870 function Effects 0.000 description 2
- 238000012067 mathematical method Methods 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000000053 physical method Methods 0.000 description 2
- 230000001133 acceleration Effects 0.000 description 1
- 230000003044 adaptive effect Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 230000007774 longterm Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 239000010865 sewage Substances 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
-
- 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
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2119/00—Details relating to the type or aim of the analysis or the optimisation
- G06F2119/02—Reliability analysis or reliability optimisation; Failure analysis, e.g. worst case scenario performance, failure mode and effects analysis [FMEA]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2119/00—Details relating to the type or aim of the analysis or the optimisation
- G06F2119/04—Ageing analysis or optimisation against ageing
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Evolutionary Computation (AREA)
- General Engineering & Computer Science (AREA)
- Artificial Intelligence (AREA)
- Computational Linguistics (AREA)
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Computer Hardware Design (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Geometry (AREA)
- Data Mining & Analysis (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention discloses an LSTM-based underground pipeline service life prediction method, which comprises the following steps: collecting strain monitoring data; preprocessing data to obtain normal time sequence data; constructing a sample set by the time sequence data, and dividing the sample set into a training set and a testing set; inputting training set data into an LSTM model for iterative training, storing the model after the training is finished, and inputting test set data into the trained LSTM model for predicting the result; and evaluating the prediction result by adopting the evaluation index. The invention adopts the LSTM neural network to predict the service life of the underground drainage pipeline. Firstly, monitoring strain data on a pipeline under the action of load; then, preprocessing the obtained monitoring data by adopting a 3 delta criterion and wavelet transformation to obtain a stable data sequence; and then training data by adopting a long-short term memory network model, and adjusting the network by using an Adam optimization algorithm to obtain an optimal prediction result, thereby achieving the purpose of predicting the service life of the network.
Description
Technical Field
The invention relates to the technical field of pipeline service life prediction, in particular to an underground pipeline service life prediction method based on LSTM.
Background
The urban underground drainage pipe network is an important component in municipal infrastructure, plays an essential role in daily life of people, and is an important guarantee for maintaining urban environment cleanness. In recent years, due to the continuous acceleration of the urbanization process, the urban impermeable pavement coverage rate is remarkably increased, and in addition, the change of the climate conditions causes frequent urban rain and flood disasters. Research shows that urban inland inundation is caused not only by extreme weather and increasing urbanization process, but also by internal reasons such as equipment failure, pipeline blockage and corrosion. As the service life of a pipeline increases, pipe sections exhibit defects such as pipe deformation, plugging, leakage, collapse, and the like. Therefore, the prediction of the service life of the drainage pipeline is very important.
The existing method for predicting the service life of the pipeline mainly comprises an electrochemical theory, a residual strength, a reliability theory, a mathematical and physical method and the like. The electrochemical theory method is simple in field test data acquisition, but is suitable for uniform corrosion in the pipeline, the residual strength is generally suitable for evaluating single defects of low-steel-grade and long-term service pipelines, the reliability strength needs comprehensive and complete data parameters to realize higher prediction precision, the existing mathematical and physical method needs various test data, and the precision of the prediction method needs to be improved. Under the condition that a pipeline is washed by sewage for a long time, various damages occur in the pipeline, so that the service life is influenced, various limitations exist in the existing method, and the prediction precision needs to be improved, so that a better method is needed for predicting the service life of the pipeline.
Disclosure of Invention
In order to solve the problems in the prior art, the invention provides the underground pipeline service life prediction method based on the LSTM, which has high prediction precision and solves the problems mentioned in the background technology.
In order to achieve the purpose, the invention provides the following technical scheme: an LSTM-based underground pipeline life prediction method comprises the following steps:
s1, collecting strain monitoring data;
s2, preprocessing data to obtain normal time sequence data;
s3, constructing a sample set by the time sequence data, and dividing the sample set into a training set and a testing set;
s4, inputting the training set data into the LSTM model for iterative training, storing the model after the training is finished, and inputting the test set data into the trained LSTM model for predicting the result;
and S5, evaluating the prediction result by adopting the evaluation index.
Preferably, the step S1 specifically includes:
s1.1, arranging strain gauges on the inner wall and the outer wall of a pipeline, wherein the strain gauges comprise sockets, spigots and sections of a pipe body;
s1.2, under the action of cyclic load, collecting strain monitoring data of each monitoring point by using a stress-strain gauge.
Preferably, the step S2 specifically includes:
s2.1, processing abnormal values in the strain monitoring data obtained in the step S1, and detecting and eliminating gross errors in the data by adopting a 3 delta criterion;
s2.2, denoising the data with the abnormal values removed by adopting wavelet transformation.
Preferably, the step S3 specifically includes:
s3.1, training and testing the set according to the following ratio of 9:1, where training set and test set are 90% and 10%, respectively;
s3.2, the training set comprises a training input trainX and a target output trainY; the test set includes test input testX and test output testY.
Preferably, the step S4 specifically includes:
s4.1, inputting the trainX and the trainY into an LSTM model for training to obtain training output
S4.2, obtaining training output through calculation of an error loss function
Difference with target output trainY;
s4.3, reversely calculating the error of each neuron, adjusting and updating the LSTM network weight parameters by using an Adam optimization algorithm, and performing iterative training on the network;
and S4.4, inputting the test set data into the trained LSTM model to predict the result.
Preferably, in step S4.3, the network is iteratively trained, and when the model is trained until the loss error of each iteration does not decrease, or the maximum number of iterations is reached, the weight parameter and the model are saved.
Preferably, the evaluation index specifically includes a root mean square error value.
Preferably, the LSTM model includes an input layer, an LSTM layer, a full connection layer, and an output layer; the training set data sequentially passes through an input layer, an LSTM layer, a full connection layer and an output layer.
The invention has the beneficial effects that: compared with the prior art, the invention adopts the LSTM neural network to predict the service life of the underground drainage pipeline, and firstly monitors the strain data on the pipeline under the load action; then, preprocessing the obtained monitoring data by adopting a 3 delta criterion and wavelet transformation to obtain a stable data sequence; and then training the data by adopting a long-short term memory network model, and adjusting the network by using an Adam optimization algorithm to obtain an optimal prediction result, thereby achieving the purpose of predicting the service life of the network.
Drawings
FIG. 1 is a flow chart of the LSTM-based underground pipe life prediction method of the present invention;
FIG. 2 is a network architecture diagram of the LSTM-based method of predicting the life of an underground pipe of the present invention;
FIG. 3 is a graph of the prediction effect of the LSTM-based underground pipeline life prediction method of 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. 1-3, the present invention provides a technical solution: an LSTM-based underground pipeline life prediction method, as shown in fig. 1, includes the following steps:
the method comprises the following steps: acquiring strain monitoring data of each position on the pipeline through a test; the specific process is as follows:
1.1, arranging strain gauges on the inner wall and the outer wall of a pipeline, wherein each strain gauge comprises eight sections of a socket, a spigot and a pipe body;
and 1.2, acquiring strain values of each monitoring point by using a stress-strain gauge under the action of cyclic load.
Step two: carrying out data preprocessing on the data acquired in the step one to obtain a normal data sequence; the specific process is as follows:
2.1, processing abnormal values in the monitoring data obtained in the step one, and mainly adopting a 3 delta criterion, namely a Lauda criterion, to detect and eliminate coarse errors in the data;
2.2, denoising the data with the abnormal values removed in the step 2.1 by adopting wavelet transformation.
Step three: constructing a sample set for the time series data obtained in the step two, wherein the sample set comprises a training input trainX, a target output trainY, a test input for verifying the validity of the model, a test output testX and a test output testY; the specific process is as follows:
3.1, dividing the sample data obtained in the step two into a training input train X, a target output train Y, a test input and output testX and testY for verifying the effectiveness of the model;
3.2 training set and test set according to 9:1, where the training and test sets are 90% and 10%, respectively.
Step four: training the training set and the test set by using the trained LSTM model, wherein the training data sequentially passes through an input layer, an LSTM layer, a full connection layer and an output layer, and fig. 2 is a structural diagram of an LSTM network; the specific process is as follows:
4.1, inputting the training sets tranX and tranY into an LSTM model for training to obtain training output
4.2 obtaining training output by calculating error loss function
And the target output trainY. The error loss function is mainly measured according to the inconsistency degree of the predicted value and the true value of the model, and the smaller the loss function is, the better the robustness of the model is.
4.3, calculating the error of each neuron reversely, designing independent adaptive learning rates for different parameters by calculating the first moment estimation and the second moment estimation of the gradient by using an Adam optimization algorithm, so as to adjust and update the LSTM network weight parameters according to the error between the target value and the predicted value, and performing iterative training on the network. When the loss error reaches a certain degree and does not drop any more when the model is trained to each iteration, or the maximum iteration number is reached, the weight parameter which enables the error of the target value and the predicted value to be minimum is found and stored.
4.4, verifying the accuracy of the prediction algorithm through the test set, and obtaining a comparison graph of a predicted value and a true value in fig. 3.
Step five: and evaluating the prediction result by taking the root mean square error as an evaluation index.
The root mean square error is expressed as:
wherein, N is the total number of samples,
to predict value, y i Are true values.
The invention adopts the LSTM neural network to predict the service life of the underground drainage pipeline. Firstly, monitoring strain data on a pipeline under the action of load; then, preprocessing the obtained monitoring data by adopting a 3 delta criterion and wavelet transformation to obtain a stable data sequence; and then training the data by adopting a long-short term memory network model, and adjusting the network by using an Adam optimization algorithm to obtain an optimal prediction result, thereby achieving the purpose of predicting the service life of the network.
Although the present invention has been described in detail with reference to the foregoing embodiments, it will be apparent to those skilled in the art that various changes in the embodiments and/or modifications of the invention can be made, and equivalents and modifications of some features of the invention can be made without departing from the spirit and scope of the invention.
Claims (8)
1. An LSTM-based underground pipeline life prediction method is characterized by comprising the following steps:
s1, collecting strain monitoring data;
s2, preprocessing data to obtain normal time sequence data;
s3, constructing a sample set by the time sequence data, and dividing the sample set into a training set and a testing set;
s4, inputting the training set data into the LSTM model for iterative training, storing the model after the training is finished, and inputting the test set data into the trained LSTM model for predicting the result;
and S5, evaluating the prediction result by adopting the evaluation index.
2. The LSTM-based underground pipe life prediction method of claim 1, wherein: the step S1 specifically includes:
s1.1, arranging strain gauges on the inner wall and the outer wall of a pipeline, wherein the strain gauges comprise sockets, spigots and sections of a pipe body;
s1.2, under the action of cyclic load, acquiring strain monitoring data of each monitoring point by using a stress-strain gauge.
3. The LSTM-based underground pipe life prediction method of claim 1, wherein: the step S2 specifically includes:
s2.1, processing abnormal values in the strain monitoring data obtained in the step S1, and detecting and eliminating gross errors in the data by adopting a 3 delta criterion;
s2.2, denoising the data with the abnormal values removed by adopting wavelet transformation.
4. An LSTM-based underground pipe life prediction method as in claim 1, wherein: the step S3 specifically includes:
s3.1, training set and test set are as follows: 1, where training set and test set are 90% and 10%, respectively;
s3.2, the training set comprises a training input trainX and a target output trainY; the test set includes test input testX and test output testY.
5. The LSTM-based underground pipe life prediction method of claim 1, wherein: the step S4 specifically includes:
s4.1, inputting the trainX and the trainY into an LSTM model for training to obtain training output
S4.2, obtaining training output through calculation of an error loss function
Difference with target output trainY;
s4.3, reversely calculating the error of each neuron, adjusting and updating the LSTM network weight parameters by using an Adam optimization algorithm, and performing iterative training on the network;
and S4.4, inputting the test set data into the trained LSTM model to predict the result.
6. The LSTM-based underground pipe life prediction method of claim 5, wherein: in step S4.3, the network is iteratively trained, and when the model is trained until the iteration loss error does not decrease any more each time, or the maximum iteration number is reached, the weight parameter and the model are saved.
7. An LSTM-based underground pipe life prediction method as in claim 1, wherein: the evaluation index specifically includes a root mean square error value.
8. A LSTM-based underground pipe life prediction method according to claim 1 or 5, wherein: the LSTM model comprises an input layer, an LSTM layer, a full connection layer and an output layer; the training set data sequentially passes through an input layer, an LSTM layer, a full connection layer and an output layer.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210729622.2A CN115310258A (en) | 2022-06-24 | 2022-06-24 | Underground pipeline service life prediction method based on LSTM |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210729622.2A CN115310258A (en) | 2022-06-24 | 2022-06-24 | Underground pipeline service life prediction method based on LSTM |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN115310258A true CN115310258A (en) | 2022-11-08 |
Family
ID=83855762
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210729622.2A Pending CN115310258A (en) | 2022-06-24 | 2022-06-24 | Underground pipeline service life prediction method based on LSTM |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115310258A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117236087A (en) * | 2023-11-16 | 2023-12-15 | 苏州顶材新材料有限公司 | Method and system for evaluating service life of thermoplastic elastomer of blending type interpenetrating network |
-
2022
- 2022-06-24 CN CN202210729622.2A patent/CN115310258A/en active Pending
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117236087A (en) * | 2023-11-16 | 2023-12-15 | 苏州顶材新材料有限公司 | Method and system for evaluating service life of thermoplastic elastomer of blending type interpenetrating network |
| CN117236087B (en) * | 2023-11-16 | 2024-01-26 | 苏州顶材新材料有限公司 | Method and system for evaluating service life of thermoplastic elastomer of blending type interpenetrating network |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111222698B (en) | Internet of things-oriented ponding water level prediction method based on long-time and short-time memory network | |
| CN113239504B (en) | Pipeline corrosion defect prediction method based on optimized neural network | |
| CN112529240B (en) | Atmospheric environment data prediction method, system, device and storage medium | |
| CN110929359B (en) | Pipe network siltation risk prediction modeling method based on PNN neural network and SWMM technology | |
| CN114547951B (en) | Dam state prediction method and system based on data assimilation | |
| CN112966856A (en) | Mountain torrent risk prediction method and prediction system | |
| CN114154226B (en) | Foundation pit stability monitoring method | |
| CN113221439B (en) | BP neural network-based drainage system real-time calibration and dynamic prediction method | |
| CN115310258A (en) | Underground pipeline service life prediction method based on LSTM | |
| CN113987908A (en) | Natural gas pipe network leakage early warning method based on machine learning method | |
| CN115654381A (en) | Water supply pipeline leakage detection method based on graph neural network | |
| CN113985496B (en) | Storm surge intelligent forecasting method based on LSTM-GM neural network model | |
| CN115049013A (en) | Short-term rainfall early warning model fusion method combining linearity and SVM | |
| Yang et al. | Comparative Study on Deformation Prediction Models of Wuqiangxi Concrete Gravity Dam Based onMonitoring Data. | |
| CN114548680A (en) | Method and system for automatically calibrating parameters of urban storm flood management model | |
| CN111553226B (en) | Method for extracting river monitoring section water surface width based on remote sensing interpretation technology | |
| CN117073933B (en) | Pipeline pressure monitoring method and system for gas pipe network | |
| CN113486295A (en) | Fourier series-based total ozone change prediction method | |
| CN116167240A (en) | Multi-measuring-point combined monitoring method for dam structure damage | |
| CN110110944A (en) | One kind predicting wind electricity digestion demand quantity algorithm based on modified wavelet neural network | |
| CN113221440B (en) | Optimal arrangement and real-time global inversion method for monitoring points of drainage system | |
| CN115062891A (en) | Method for quickly positioning leakage of long-distance pressure water delivery pipeline | |
| CN110705152B (en) | Acceleration sensor arrangement method for monitoring earthquake resistance of building electromechanical pipeline | |
| CN117520996B (en) | Real-time pressure testing method for culvert pipe | |
| CN115358475B (en) | Disaster prediction method and system based on support vector machine and gray BP neural network |
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 |