CN117009788A - Buried fluid delivery pipeline perimeter collapse early warning method, storage medium and method based on water hammer characteristic parameter set - Google Patents
Buried fluid delivery pipeline perimeter collapse early warning method, storage medium and method based on water hammer characteristic parameter set Download PDFInfo
- Publication number
- CN117009788A CN117009788A CN202311079873.1A CN202311079873A CN117009788A CN 117009788 A CN117009788 A CN 117009788A CN 202311079873 A CN202311079873 A CN 202311079873A CN 117009788 A CN117009788 A CN 117009788A
- Authority
- CN
- China
- Prior art keywords
- collapse
- water hammer
- neural network
- pipeline
- training
- 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 81
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 title claims abstract description 68
- 239000012530 fluid Substances 0.000 title claims abstract description 11
- 238000003860 storage Methods 0.000 title claims abstract description 11
- 238000013528 artificial neural network Methods 0.000 claims abstract description 68
- 238000012549 training Methods 0.000 claims abstract description 59
- 238000004422 calculation algorithm Methods 0.000 claims abstract description 31
- 238000010801 machine learning Methods 0.000 claims abstract description 20
- 238000012360 testing method Methods 0.000 claims abstract description 13
- 238000004458 analytical method Methods 0.000 claims abstract description 8
- 238000005516 engineering process Methods 0.000 claims abstract description 6
- 230000006870 function Effects 0.000 claims description 78
- 239000003016 pheromone Substances 0.000 claims description 38
- 238000005457 optimization Methods 0.000 claims description 37
- 239000011159 matrix material Substances 0.000 claims description 28
- 239000000523 sample Substances 0.000 claims description 18
- 239000013598 vector Substances 0.000 claims description 15
- 241000257303 Hymenoptera Species 0.000 claims description 14
- 230000008569 process Effects 0.000 claims description 13
- 238000007637 random forest analysis Methods 0.000 claims description 13
- 238000012706 support-vector machine Methods 0.000 claims description 11
- 238000009826 distribution Methods 0.000 claims description 10
- 238000007781 pre-processing Methods 0.000 claims description 10
- 230000005540 biological transmission Effects 0.000 claims description 8
- 238000004088 simulation Methods 0.000 claims description 8
- 238000012795 verification Methods 0.000 claims description 8
- 238000003062 neural network model Methods 0.000 claims description 7
- 239000000725 suspension Substances 0.000 claims description 7
- 230000001133 acceleration Effects 0.000 claims description 6
- 238000002474 experimental method Methods 0.000 claims description 6
- 230000007547 defect Effects 0.000 claims description 5
- 230000014509 gene expression Effects 0.000 claims description 5
- 238000004590 computer program Methods 0.000 claims description 4
- 239000002184 metal Substances 0.000 claims description 4
- 230000002093 peripheral effect Effects 0.000 claims description 4
- 230000009471 action Effects 0.000 claims description 3
- 238000013135 deep learning Methods 0.000 claims description 3
- 238000002360 preparation method Methods 0.000 claims description 3
- 230000011218 segmentation Effects 0.000 claims description 3
- 230000003595 spectral effect Effects 0.000 claims description 3
- 238000012423 maintenance Methods 0.000 abstract description 13
- 238000012544 monitoring process Methods 0.000 abstract description 11
- 230000000694 effects Effects 0.000 abstract description 9
- 238000000605 extraction Methods 0.000 abstract description 3
- 230000009194 climbing Effects 0.000 abstract 1
- 238000010276 construction Methods 0.000 description 5
- 238000011156 evaluation Methods 0.000 description 5
- 230000009467 reduction Effects 0.000 description 4
- 238000001228 spectrum Methods 0.000 description 4
- 238000001514 detection method Methods 0.000 description 3
- 238000011161 development Methods 0.000 description 3
- 230000018109 developmental process Effects 0.000 description 3
- 238000001914 filtration Methods 0.000 description 3
- 230000007246 mechanism Effects 0.000 description 3
- 238000010183 spectrum analysis Methods 0.000 description 3
- 229910000831 Steel Inorganic materials 0.000 description 2
- 230000008901 benefit Effects 0.000 description 2
- 230000008859 change Effects 0.000 description 2
- 238000002790 cross-validation Methods 0.000 description 2
- 238000012545 processing Methods 0.000 description 2
- 238000005070 sampling Methods 0.000 description 2
- 230000035945 sensitivity Effects 0.000 description 2
- 239000002689 soil Substances 0.000 description 2
- 239000010959 steel Substances 0.000 description 2
- 230000004913 activation Effects 0.000 description 1
- 230000002238 attenuated effect Effects 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 238000010168 coupling process Methods 0.000 description 1
- 238000005859 coupling reaction Methods 0.000 description 1
- 238000007405 data analysis Methods 0.000 description 1
- 238000000354 decomposition reaction Methods 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 238000005265 energy consumption Methods 0.000 description 1
- 230000005284 excitation Effects 0.000 description 1
- 239000000284 extract Substances 0.000 description 1
- 230000036541 health Effects 0.000 description 1
- 238000011423 initialization method Methods 0.000 description 1
- 238000007689 inspection Methods 0.000 description 1
- 239000000463 material Substances 0.000 description 1
- 210000002569 neuron Anatomy 0.000 description 1
- 230000002265 prevention Effects 0.000 description 1
- 230000001902 propagating effect Effects 0.000 description 1
- 230000008707 rearrangement Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 230000000717 retained effect Effects 0.000 description 1
- 239000010865 sewage Substances 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 238000012800 visualization Methods 0.000 description 1
- 239000002699 waste material Substances 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/213—Feature extraction, e.g. by transforming the feature space; Summarisation; Mappings, e.g. subspace methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/004—Artificial life, i.e. computing arrangements simulating life
- G06N3/006—Artificial life, i.e. computing arrangements simulating life based on simulated virtual individual or collective life forms, e.g. social simulations or particle swarm optimisation [PSO]
-
- 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
- G06N3/0464—Convolutional networks [CNN, ConvNet]
-
- 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)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- General Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Evolutionary Computation (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- Software Systems (AREA)
- Mathematical Physics (AREA)
- Health & Medical Sciences (AREA)
- Biomedical Technology (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Computing Systems (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Evolutionary Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Pipeline Systems (AREA)
Abstract
The application discloses a buried fluid delivery pipeline perimeter collapse early warning method, a storage medium and a method based on a water hammer characteristic parameter set, which are characterized in that a large amount of data are collected by analyzing the vibration characteristics of a valve closing water hammer, the time domain and frequency domain sensitive characteristics of vibration signals are extracted and analyzed, a plurality of machine learning algorithms are used for preliminary training and test analysis, a radial basis neural network machine learning method with the best recognition effect is selected, and model parameters are optimized by adopting an ant colony algorithm; further, the original data set is subjected to extraction and reconstruction on the original signals by adopting a wavelet packet technology, and characteristic information parameters of the extracted sensitive frequency band are combined with the original parameters, so that the model precision is remarkably improved, and the final recognition accuracy is close to 98%. The method system not only can be used for safety monitoring of a pipeline system, but also can be integrated with equipment such as a climbing robot and the like, and realizes real-time monitoring and maintenance of key equipment such as an electromagnetic valve and the like. Thereby providing effective guarantee for the pipeline transportation safety.
Description
Technical Field
The application relates to the technical field of pipeline defect detection, in particular to a buried pipeline perimeter collapse early warning method, a storage medium and a method based on a water hammer characteristic parameter set.
Background
Currently, buried pipeline collapse is a common and serious problem in urban infrastructure operations and maintenance. When the pipeline structure is damaged or the soil stability is changed, the pipeline may collapse or break, so that basic services such as water supply, air supply, sewage treatment and the like are interrupted, and great inconvenience and economic loss are brought to the life and work of people. Because buried pipelines are usually located underground, detection and monitoring of the buried pipelines are relatively difficult, and traditional inspection and maintenance modes often cannot discover changes and potential risks of pipeline health conditions in time. Therefore, the development and application of the intelligent pipeline monitoring system in the patent background technology has important background significance:
the safety is improved: the intelligent pipeline monitoring system can discover pipeline structure problems and soil changes as early as possible through real-time monitoring and data analysis and provide early warning and advice, so that maintenance measures are taken timely, the risk of pipeline collapse and breakage is reduced, and the safety of the public is guaranteed.
Reliability is improved: through accurate evaluation and prediction of the intelligent pipeline monitoring system, operation and maintenance personnel can make a more effective maintenance plan, optimize maintenance work, reduce the occurrence probability of pipeline faults and improve the reliability of basic services such as water supply, air supply and the like.
The cost is reduced: traditional pipeline maintenance is often periodically patrolled and inspected or remedied afterwards, and is low in efficiency and high in maintenance cost. The intelligent pipeline monitoring system can realize remote monitoring and accurate prediction, helps to optimize maintenance plans, and reduces unnecessary maintenance frequency and maintenance range, thereby reducing maintenance cost.
Promote sustainable development: the intelligent pipeline monitoring system can improve the service life of the pipeline, reduce the construction requirement on a new pipeline, reduce resource waste and energy consumption, and promote sustainable development of urban infrastructure. None of the prior art discloses a technique related to the above-mentioned research.
Disclosure of Invention
The application aims to provide a buried pipeline perimeter collapse early warning method, a storage medium and a method based on a water hammer characteristic parameter set, which can accurately identify early pipeline collapse types, thereby providing assistance for pipeline collapse early warning.
The application adopts the technical scheme that:
a buried pipeline perimeter collapse early warning method based on a water hammer characteristic parameter set comprises the following steps:
s1, uniformly defining and classifying defects by utilizing different peripheral collapse degrees of the buried pipeline, and dividing the early collapse of the buried pipeline into three conditions of bottom collapse, exposed surface and complete suspension; then collecting vibration data of the experiment platform pipeline system by using an acceleration sensor; collecting axial vibration data of a water hammer, wherein the axial direction is the extending direction of a pipeline;
s2, carrying out data batch preprocessing on vibration data collected by a pipeline system to obtain sample data, wherein denoising and data segmentation are adopted in the preprocessing;
s3, extracting features: extracting time domain sensitive characteristic information and frequency domain characteristics of water hammer vibration signals caused by early collapse of different buried pipelines in the preprocessed sample data to form a characteristic set; the time domain sensitive characteristic information is obtained by establishing a finite element simulation model by COMSOL, simulating valve closing action by setting a valve closing function to enable a pipeline to generate water hammer vibration waves, and carrying out pickup analysis on the water hammer vibration wave information by a point probe;
s4, according to the result of the COMSOL simulation model, closing the electromagnetic valve mainly to generate a water hammer and enabling the electromagnetic valve metal ball to hit the pipeline to generate vibration waves, and extracting vibration signal characteristics which are sensitive to the early collapse type of the perimeter of the flow transmission pipeline in the vicinity of 450 HZ;
s5, constructing different machine learning basic models, and training a support vector machine, a random forest, a radial basis neural network and a BP neural network; cross verification is adopted to verify the accuracy of the model, so that a machine learning model which is most suitable for further optimization is selected;
and S6, further optimizing the radial basis function neural network model, namely optimizing the parameter setting of the radial basis function neural network mainly through an ant colony algorithm.
S7, training the optimized machine learning model to obtain a parameter set with the highest recognition rate;
and S8, testing the parameter set with the highest recognition rate obtained in the step S7, and respectively recognizing early collapse to obtain the type of the early collapse of the perimeter of the flow transmission pipeline.
The time domain characteristic parameters and expressions extracted in the step S3 are as follows:
wherein x (N) represents a time domain sequence of the signal, n=1, 2, …, N and N are the number of sample points, the signal is described by an average value A1, and the energy of the water hammer vibration signal under different conditions can be detected; the variance A2 represents the dynamic component of the signal energy, reflects the discrete degree between data, and has better accuracy in the aspects of model prediction and experimental data description; the effective value A3 describes the energy of the vibration signal;
the specific characteristic frequency domain parameter formula is as follows:
wherein S (K) represents a frequency domain sequence of the signal, where k=1, 2, …, K is the number of sample points; frequency domain feature statistics parameters are introduced to describe the characteristics of the signal: the spectral mean B1 describes its fluctuations, and the frequency distribution of the pipe vibration signal is described by the average frequency B2 and the root mean square frequency B3;representing the square of the corresponding frequency domain sequence.
The feature set in step S3 includes: bottom collapse, exposed surface and complete suspension, and time and frequency domain feature sets of the three conditions.
In step S4, when the characteristic of the sensitive vibration signal is extracted, the signal is decomposed and reconstructed by adopting a wave packet technology to obtain a desired wavelet packet frequency band, and then the frequency domain characteristic parameter is extracted; the specific wavelet packet selects three layers of wavelets, which are divided into 8 frequency bands, and Haar is a wavelet basis function.
In step S5, referring to MATLAB deep learning tool library to perform preliminary training test, selecting the most suitable machine learning method to perform test optimization, wherein the main steps are as follows:
data preparation: training data are arranged into a feature matrix X and a corresponding label vector Y; ensuring that the data has been standardized or preprocessed;
support vector machine training: creating a support vector machine model object using the fitcvm function; training a model by using a train method, and transmitting a feature matrix X and a label vector Y; the new data can be classified and predicted by using a prediction method;
random forest training: creating a random forest object by using the TreeBagger function; training a random forest model by using a train method, and transmitting a feature matrix X and a label vector Y; the new data can be classified and predicted by using a prediction method;
radial basis function neural network training: creating a radial basis function neural network object using the newgrnn function; training a network model by using a train method, and transmitting a feature matrix X and a label vector Y; the sim method can be used for carrying out classified prediction on new data;
BP neural network training: creating a BP neural network object by using a feedback forwarding net function; setting a network structure and training parameters; training a network model by using a train method, and transmitting a feature matrix X and a label vector Y; classification predictions can be made for new data using sim methods.
In step S6, the radial basis function neural network structure mainly comprises an input layer, a radial base layer and an output layer; the network structure relation is mainly adjusted according to the training set, and then the center point and the width parameters are determined.
The step S6 specifically comprises the following steps:
s61: collecting and collating pipeline data for training and testing,
s62: preprocessing data, and extracting various characteristic values;
s63: establishing a basic radial basis function neural network, including initialization of the structure, weight and bias of the network and radial basis functions, S64: iterative optimization is carried out on the hidden layer and the super parameters of the radial base network through an ant colony algorithm,
s65: determining a network structure;
s66: further optimizing parameters by a backprojection algorithm.
The step S64 specifically includes the following steps:
initializing parameters of an ant colony algorithm, including the number of ants, the volatilization rate of pheromones and the initial concentration of the pheromones;
setting a proper numerical value according to actual conditions to perform ant iterative optimization;
the specific process is as follows:
s64.1 initializing ant colony parameters: setting the number and initial positions of ants, determining the scale of an ant colony, and distributing an initial position for each ant randomly or according to a specific strategy; the weight and bias of each ant are randomly initialized and used as the initial parameters of the neural network;
s64.2 initializing a pheromone matrix: creating an pheromone matrix corresponding to the problem scale, wherein the initial value can be set to be the same constant or can be adjusted according to the characteristics of the problem;
s64.3 routing of ants: each ant carries out forward propagation on a training data set according to the current neural network parameters, calculates the value of a loss function, and updates the neural network parameters according to the pheromone concentration and the heuristic function;
s64.4 pheromone update: updating the pheromone concentration matrix according to the path quality of each ant; a better path will increase the pheromone concentration and a worse path will decrease the pheromone concentration;
s64.5, repeatedly performing iterative optimization, and judging termination conditions: setting iteration times or stopping conditions: setting proper iteration times or stopping conditions according to practical problems and experience, for example, reaching the maximum iteration times and meeting convergence conditions; judging whether a termination condition is met;
s64.6 updating the neural network parameters: in the iteration process, recording and storing the optimal solution in each iteration, and updating the weight and bias of the neural network according to the neural network parameters corresponding to the optimal path;
s64.7, returning to the neural network model subjected to the ant colony algorithm depth optimization.
A computer readable storage medium having stored thereon a computer program which, when executed by a processor, causes a device in which the computer readable storage medium is located to perform the buried fluid transport pipeline perimeter collapse warning method based on a water hammer characteristic parameter set according to any one of claims 1 to 8.
An electronic device comprising a processor and a memory, wherein the memory stores a program which can be run on the processor, the processor realizes the buried pipeline perimeter collapse early warning method based on the machine learning of the water hammer characteristic parameter set according to any one of 1-8 when executing the program,
the method has the advantages that the classification effect of the radial basis function neural network algorithm model subjected to the parameter feature set depth optimization on the sample data is very remarkable, the effective discrimination of the early collapse type can be realized, the identification precision of the optimized radial basis function neural network can reach 97%, and meanwhile, the effectiveness of the selected feature parameter set on the identification of the collapse type is verified by the ant colony algorithm depth optimization of the radial basis function neural network. And finally, verifying the effectiveness of the selected characteristic parameters for identifying the early collapse type by using an ant colony algorithm depth optimization radial basis function neural network method. The method of the application can collect different sample data under different working scenes, different equipment, different materials, and the like, but the whole type prediction method is applicable to all occasions.
Drawings
In order to more clearly illustrate the embodiments of the application or the technical solutions in the prior art, the drawings that are required in the embodiments or the description of the prior art will be briefly described, it being obvious that the drawings in the following description are only some embodiments of the application, and that other drawings may be obtained according to these drawings without inventive effort for a person skilled in the art.
FIG. 1 is a flow chart of the present application.
FIG. 2 is a graph showing the variation of the water hammer pressure according to the present application under various conditions.
FIG. 3 is a graph showing the third-order characteristic frequency variation according to the present application.
FIG. 4 is a diagram of the original signal and the specific frequency band according to the present application
FIG. 5 is a two-dimensional scatter plot of feature parameter visualization in accordance with the present application.
Fig. 6 is a flowchart of the ant colony algorithm optimization radial basis function neural network according to the present application.
FIG. 7 is a graph of the effect of the radial basis function neural network before and after optimization on early collapse type identification according to the present application.
Detailed Description
The following description of the embodiments of the present application will be made clearly and completely with reference to the accompanying drawings, in which it is apparent that the embodiments described are only some embodiments of the present application, but not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the application without any inventive effort, are intended to be within the scope of the application.
As shown in fig. 1, the present application includes the steps of:
s1, uniformly defining and classifying defects by utilizing different peripheral collapse degrees of the buried pipeline, and dividing the early collapse of the buried pipeline into three conditions of bottom collapse, exposed surface and complete suspension; then collecting vibration data of the experiment platform pipeline system by using an acceleration sensor; collecting axial vibration data of a water hammer, wherein the axial direction is the extending direction of a pipeline;
s2, carrying out data batch preprocessing on vibration data collected by a pipeline system to obtain sample data, wherein denoising and data segmentation are adopted in the preprocessing;
s3, extracting features: extracting time domain sensitive characteristic information and frequency domain characteristics of water hammer vibration signals caused by early collapse of different buried pipelines in the preprocessed sample data to form a characteristic set; the time domain sensitive characteristic information is obtained by establishing a finite element simulation model by COMSOL, simulating valve closing action by setting a valve closing function to enable a pipeline to generate water hammer vibration waves, and carrying out pickup analysis on the water hammer vibration wave information by a point probe; the time domain characteristic parameters and expressions extracted in the step S3 are as follows:
wherein x (N) represents a time domain sequence of the signal, n=1, 2, …, N and N are the number of sample points, the signal is described by an average value A1, and the energy of the water hammer vibration signal under different conditions can be detected; the variance A2 represents the dynamic component of the signal energy, reflects the discrete degree between data, and has better accuracy in the aspects of model prediction and experimental data description; the effective value A3 describes the energy of the vibration signal;
the specific characteristic frequency domain parameter formula is as follows:
wherein S (K) represents a frequency domain sequence of the signal, where k=1, 2, …, K is the number of sample points; frequency domain feature statistics parameters are introduced to describe the characteristics of the signal: the spectral mean B1 describes its fluctuations, and the frequency distribution of the pipe vibration signal is described by the average frequency B2 and the root mean square frequency B3; wherein the average frequency B2 is the average value of the vibration frequency of the pipeline, the root mean square frequency B3 is the arithmetic square root of the mean square frequency, and the mean square frequency is the weighted average of the square of the signal frequency.
The feature set in step S3 includes: bottom collapse, exposed surface and complete suspension, and time and frequency domain feature sets of the three conditions.
S4, according to the result of the COMSOL simulation model, closing the electromagnetic valve mainly to generate a water hammer and enabling the electromagnetic valve metal ball to hit the pipeline to generate vibration waves, and extracting vibration signal characteristics which are sensitive to the early collapse type of the perimeter of the flow transmission pipeline in the vicinity of 450 HZ; in step S4, when the characteristic of the sensitive vibration signal is extracted, the signal is decomposed and reconstructed by adopting a wave packet technology to obtain a desired wavelet packet frequency band, and then the frequency domain characteristic parameter is extracted; the specific wavelet packet selects three layers of wavelets, which are divided into 8 frequency bands, and Haar is a wavelet basis function.
S5, constructing different machine learning basic models, and training a support vector machine, a random forest, a radial basis neural network and a BP neural network; cross verification is adopted to verify the accuracy of the model, so that a machine learning model which is most suitable for further optimization is selected;
and S6, further optimizing the radial basis function neural network model, namely optimizing the parameter setting of the radial basis function neural network mainly through an ant colony algorithm. In step S6, the radial basis function neural network structure mainly comprises an input layer, a radial base layer and an output layer; the network structure relation is mainly adjusted according to the training set, and then the center point and the width parameters are determined.
The step S6 specifically comprises the following steps:
s61: collecting and collating pipeline data for training and testing,
s62: preprocessing data, and extracting various characteristic values;
s63: establishing a basic radial basis function neural network, including initialization of the structure, weight and bias of the network and radial basis functions, S64: iterative optimization is carried out on the hidden layer and the super parameters of the radial base network through an ant colony algorithm,
s65: determining network structure
S66: further optimizing parameters by a backprojection algorithm.
The step S64 specifically includes the following steps:
initializing parameters of an ant colony algorithm, including the number of ants, the volatilization rate of pheromones and the initial concentration of the pheromones;
setting a proper numerical value according to actual conditions to perform ant iterative optimization;
the specific process is as follows:
s64.1 initializing ant colony parameters: setting the number and initial positions of ants, determining the scale of an ant colony, and distributing an initial position for each ant randomly or according to a specific strategy; the weight and bias of each ant are randomly initialized and used as the initial parameters of the neural network;
s64.2 initializing a pheromone matrix: creating an pheromone matrix corresponding to the problem scale, wherein the initial value can be set to be the same constant or can be adjusted according to the characteristics of the problem;
s64.3 routing of ants: each ant carries out forward propagation on a training data set according to the current neural network parameters, calculates the value of a loss function, and updates the neural network parameters according to the pheromone concentration and the heuristic function;
s64.4 pheromone update: updating the pheromone concentration matrix according to the path quality of each ant; a better path will increase the pheromone concentration and a worse path will decrease the pheromone concentration;
s64.5, repeatedly performing iterative optimization, and judging termination conditions: setting iteration times or stopping conditions: setting proper iteration times or stopping conditions according to practical problems and experience, for example, reaching the maximum iteration times and meeting convergence conditions; judging whether a termination condition is met;
s64.6 updating the neural network parameters: in the iteration process, recording and storing the optimal solution in each iteration, and updating the weight and bias of the neural network according to the neural network parameters corresponding to the optimal path;
s64.7, returning to the neural network model subjected to the ant colony algorithm depth optimization.
S7, training the optimized machine learning model to obtain a parameter set with the highest recognition rate;
and S8, testing the parameter set with the highest recognition rate obtained in the step S7, and respectively recognizing early collapse to obtain the type of the early collapse of the perimeter of the flow transmission pipeline.
A computer readable storage medium, on which a computer program is stored, which when executed by a processor, causes a device in which the computer readable storage medium is located to perform the above-described buried pipeline perimeter collapse warning method based on a water hammer characteristic parameter set.
An electronic device comprising a processor and a memory, wherein the memory stores a program capable of running on the processor, and the processor realizes the buried pipeline perimeter collapse early warning method based on the machine learning of the water hammer characteristic parameter set when executing the program.
The application uses the ant colony algorithm to deeply optimize the radial basis function neural network, optimizes the related parameters through the ant colony algorithm, and has the optimization capability of the ant colony algorithm more advantageous in the currently used main stream optimization algorithm. And by means of cross verification, iterative optimization is firstly carried out on the number of hidden layer nodes in the setting range of the super-parameters and the empirical formula, and then iterative optimization is carried out on the initial weight and the offset, so that the training capacity of the radial basis function neural network is enabled to achieve an optimal effect. When cross-validation is used to optimize the radial basis function neural network, the overall optimization problem is decomposed into the following steps: dividing a data set, selecting parameters, training a model, evaluating the model, adjusting the super parameters, repeating the steps until the parameter combination with the best performance is found, and obtaining a radial basis function neural network model with better performance. The method using cross validation can help solve the problem of overall optimization of the radial basis function neural network, and improve the accuracy and generalization capability of the model.
Further, for a further explanation of the method of the present application in detail, a detailed process description is given below with specific examples.
S1, a simple pipeline model is built through COMSOL simulation, as shown in fig. 2, a sigmoid function is adopted for setting a simulated valve closing function of the fluid flow velocity of the pipeline outlet, fluid inertia is utilized to generate a water hammer, the water hammer is collapsed at the bottom, the surface is exposed and is completely suspended through comparison of transmission characteristics of the three conditions, and the sensitivity characteristics of time domain signals and characteristic frequencies are shown in fig. 2. The peak attenuation influence of the water hammer is larger in the time domain, and the full-coupling characteristic frequency is solved on the basis of the geometric model. The third and fourth order characteristic frequencies are relatively sensitive. Based on these rules, the sensitive characteristic parameters of the valve closing pipe vibration are analyzed, as shown in fig. 3.
The pipeline experiment verification mainly comprises three parts of a water hammer device, an experiment data acquisition system and a pipe Zhou Niantu, wherein the water hammer device comprises a DN20 pipeline, an automatic electromagnetic valve, a flowmeter and other system components, the electromagnetic valve can be quickly opened and closed to generate water hammer excitation, and meanwhile, the water flow speed is controlled to be 0.5m/s. The pipeline system is supported by the cantilever and restrained on the steel frame, so that the contact interface of the pipe and the steel frame is ensured not to influence the experiment. The rubber box accommodates a part of the pipeline as a buried pipeline scene, so that the early collapse condition of different buried pipelines can be simulated. The transmission characteristics of the water hammer vibration signals under the three conditions that the pipeline is collapsed at the bottom, the surface is exposed and is completely suspended are studied, and experimental data are acquired through a data acquisition system for analysis. The results are shown in the upper part of fig. 4, where the acceleration sensor is used to collect data for analysis.
According to the practical conditions of the pipeline, defects are uniformly defined and classified by utilizing the difference of the peripheral collapse degree of the buried pipeline, and the early collapse of the buried pipeline is divided into three conditions of bottom collapse, exposed surface and complete suspension. And then establishing a three-dimensional finite element model of the buried pipeline, wherein each early collapse is uniformly distributed in a defined relative position in the three-dimensional finite element model, and simulating the early collapse of the pipeline by utilizing COMSOL to obtain corresponding water hammer signals, so as to obtain the water hammer vibration signals under different collapse types.
S2, preprocessing data, namely firstly carrying out noise reduction operation on the collected pipeline vibration acceleration signals, wherein the following steps are general:
importing data: the pipeline vibration acceleration signal is imported into the MATLAB environment, and an importdata or load function can be used.
High-pass filtering: if low frequency noise is present in the signal, high pass filtering may be used to filter out. The high pass filter may be designed using a highpass function or a design filter function of the signal processing tool library and applied to the signal using a filter function.
Denoising: various noise reduction methods are provided in signal processing kits, one of which is commonly used is wavelet transform noise reduction. The following is the basic step of noise reduction using wavelet transforms:
a. wavelet decomposition is performed on the signal using wavedec functions to obtain wavelet coefficients.
b. The wavelet coefficients are thresholded, the wavelet coefficients of low amplitude are set to zero, and the wavelet coefficients of high amplitude are retained.
c. The processed wavelet coefficients are reconstructed into a denoised signal using a waverec function.
The sampling point set by the data acquisition system is 3000, and only signals smaller than 1500HZ can be acquired according to the sampling theorem, so that low-pass filtering is not needed.
S3, for the time domain and frequency domain feature extraction process of the vibration signal, the feature and the expression in the following table are set into a feature extraction function through MATLAB to be directly referenced.
TABLE 1
In the time domain parameters, the average value is the average value of a group of data, and the overall average level of the system can be known by calculating the average value of pressure or flow data in a period of time, so that the water hammer phenomenon is primarily estimated. Variance is a measure of the degree of discretization of a set of data, describing fluctuations in the data. In water hammer problems, the variance may reflect the instability of the system and the strength of the pressure fluctuations. The effective value is the mean root-mean-square of a set of data sums, also known as the root-mean-square value. The effective value can measure the energy of the water hammer phenomenon, namely the amplitude of pressure or flow.
By means of the average value, the average state of the system can be known, and whether the overall level is higher or lower can be judged. The variance can help us evaluate the stability and fluctuation of the system and understand the strength of the water hammer. The effective value is an index for measuring the energy of the water hammer and can be used for evaluating the severity of the water hammer and the influence on the system.
In the frequency domain parameters, the spectrum mean value is calculated by carrying out spectrum analysis on the pressure data. The spectrum mean value can reflect the distribution condition of the water hammer phenomenon on the frequency domain. By observing the change of the spectrum mean value, the frequency range and the main frequency component of the water hammer phenomenon can be judged. The mean square frequency refers to the frequency point of the energy concentration calculated by spectrum analysis. The mean square frequency may reflect the dominant vibration frequency of the water hammer phenomenon in the frequency domain. By monitoring the change of the mean square frequency, the frequency characteristic of the water hammer phenomenon and the possible resonance condition can be judged. The root mean square frequency is the square sum root opening number of the frequency distribution obtained through frequency spectrum analysis and calculation, and is also an index for reflecting the energy distribution condition of the water hammer phenomenon on the frequency domain. The root mean square frequency can quantify the energy of the water hammer phenomenon at different frequencies, and help to know the frequency characteristics and the energy distribution of the water hammer.
Through the spectrum mean value, the distribution of the pipeline valve closing vibration on different frequencies can be known, and possible frequency components are found.
The mean square frequency can help us determine the dominant vibration frequency, judging whether there is a risk of frequency resonance. The root mean square frequency can quantify the energy distribution in the frequency domain, helping us to know the frequency characteristics and fluctuation of the system. By analyzing the frequency domain characteristic parameters, the characteristics in the frequency domain can be more comprehensively known, and the conditions of frequency distribution and energy concentration are revealed. The application extracts 6 identification parameters of the expression collapse types, namely parameters A1, B1, A2, B2, A3 and B3, and in the embodiment, the signals are decomposed and reconstructed by utilizing a wavelet packet technology to obtain a desired wavelet packet frequency band, and then the frequency domain characteristic parameters are extracted. Further, through a MATLAB deep learning tool library, the collected feature set is subjected to a preliminary training test, the advantages and disadvantages of various machine learning methods are analyzed, and a radial basis neural network with the best effect is selected for further optimization and parameter adjustment.
S4, in order to further improve the training effect, the training parameter set is optimized first, according to the COMSOL simulation model result, except for the water hammer, the electromagnetic valve is closed, resonance is generated when the electromagnetic valve metal ball hits the pipeline, and according to FIG. 3, the third resonance point is sensitive to the early collapse type. A new combination is proposed, namely, vibration signals of a specific frequency band are extracted and analyzed to obtain frequency domain characteristics, and as shown in fig. 4, vibration signals of a pipeline are extracted to obtain sensitive frequency bands to obtain frequency domain characteristics. The new feature set is learned through radial basis function neural network training.
S5, training is carried out by a support vector machine, a random forest, a radial basis function neural network and a BP neural network. And (5) adopting cross verification to verify the accuracy of the model. The purpose is to select the best fit for further optimization i The main steps of the machine learning model are as follows:
data preparation: training data are organized into a feature matrix X and corresponding label vectors Y. Ensuring that the data has been standardized or preprocessed.
Support Vector Machine (SVM) training: a. a support vector machine model object is created using the fitsvm function. b. Training a model by using the train method, and transmitting a feature matrix X and a label vector Y. c. The new data can be classified and predicted by using the prediction method.
Random forest training: a. a random forest object is created using TreeBagger functions. b. Training a random forest model by using a train method, and transmitting a feature matrix X and a label vector Y. c. The new data can be classified and predicted by using the prediction method.
Radial basis function neural network training: a. a radial basis function object is created using the newgrnn function. b. Training a network model by using the train method, and transmitting a feature matrix X and a label vector Y. c. Classification predictions can be made for new data using sim methods.
BP neural network training: a. a BP neural network object is created using the feedforwardnet function. b. Setting the structure of the network and training parameters. c. Training a network model by using the train method, and transmitting a feature matrix X and a label vector Y. d. Classification predictions can be made for new data using sim methods.
The radial basis function neural network effect can be obtained from the comprehensive accuracy of table 2 to be above the other three.
TABLE 2
S6, according to the recognition and optimization method of the characteristic parameter set of the early perimeter collapse vibration information of the buried fluid delivery pipeline based on machine learning of the S5, the method is characterized in that in the step S5, in the training process of the radial basis function neural network, data are preprocessed: including data cleansing, feature selection, etc., to improve the quality and usability of the data. Construction of a neural network: the network structure is determined, including the number of neurons of the input layer, hidden layer, and output layer, etc. At the same time, the appropriate activation function and loss function are selected. Initialization of weights and biases: the weights and biases in the neural network are initialized, and generally a random initialization method can be used. Optimization of parameters was performed using a back propagation algorithm (backprojection): the loss function is minimized by propagating the input data in the forward direction, then calculating the loss and updating the weights and biases using a back propagation algorithm. Iterative training: step 4 is repeated until a stopping condition is met, such as a maximum number of iterations or loss function convergence is reached. Model evaluation and optimization: and evaluating the model obtained through training by using a verification set or a cross verification method and the like, and optimizing the model according to an evaluation result, such as adjusting a network structure, adjusting a learning rate and the like.
The optimization process requiring manual experience to adjust can be automatically and deeply optimized through an ant colony algorithm, and the classification accuracy and convergence speed of the neural network can be improved, so that the early collapse recognition effect of the buried pipeline is improved, the optimization process is shown in fig. 6, and the specific steps are as follows:
initializing parameters: setting the number and initial positions of ants: the ant colony size is determined and each ant is assigned an initial position either randomly or according to a specific strategy.
Initializing a pheromone matrix: an pheromone matrix corresponding to the problem size is created, and the initial value can be set to be the same constant or adjusted according to the characteristics of the problem.
The moving mechanism of ants: the next moving position is selected according to a certain probability. The probability of selection may be determined based on the pheromone concentration and heuristic factors using a roulette selection mechanism.
Construction of solutions: ants construct solutions based on the chosen location, typically by traversal or path join.
Updating the pheromone: pheromone updating mechanism: updating the pheromone matrix according to the construction solution of ants and the evaluation function of the problems. A common update strategy is to update the pheromone matrix by the pheromone increment on the track of each movement of the ant.
Volatilizing pheromone: after each update, the pheromone is attenuated at a certain volatilization rate to increase the global and exploratory capabilities of the search.
Repeating the iteration: setting iteration times or stopping conditions: and setting proper iteration times or stopping conditions according to practical problems and experiences, such as reaching the maximum iteration times, meeting convergence conditions and the like.
Iterative optimization: in each iteration, the movement and construction of the solution of the ants, and the updating of the pheromones, are continuously updated through the previous steps until the stop condition is reached.
Outputting and storing an optimal solution: in the iteration process, the optimal solution in each round of iteration is recorded and stored.
Outputting a result: and after the algorithm is finished, determining a global optimal solution according to the optimal solution and an evaluation criterion.
Each early collapse type was classified into 1250 as a training set and 350 as a test set. And 6, the number of nodes of an input layer of the radial basis function neural network is 3, the number of nodes of an output layer is 12 after optimizing the number of nodes of an hidden layer, and the output result of the radial basis function neural network is rounded. In this embodiment, the bottom is collapsed, the surface is exposed, and the data set class labels in these three cases are set to 1,2, and 3, respectively.
As can be seen from fig. 7, the optimized radial basis function neural network algorithm model has a significantly improved classification effect on sample data, can realize effective discrimination of early collapse types, has an identification accuracy of 98% after optimization, and simultaneously, the validity of the selected identification parameters for identifying early collapse types is verified by optimizing the radial basis function neural network through the ant colony algorithm.
The application provides a buried pipeline perimeter collapse early warning method based on water hammer wave characteristic parameter set machine learning, which can accurately identify the type of the buried pipeline perimeter early collapse and has great engineering significance and good application prospect.
Further, the sensitivity characteristic analysis of closing the water hammer vibration and early collapse of the periphery of the buried pipeline through the COMSOL analysis electromagnetic valve is performed, corresponding identification characteristic parameters and frequency domain characteristics of sensitive frequency bands are verified and extracted through an experimental platform, the effectiveness of different characteristic parameter sets on inversion of the early collapse type of the periphery of the buried pipeline is analyzed, convenience and accuracy of pipeline collapse identification can be improved in the field of buried pipeline detection, and the method is significant for collapse prevention of the buried pipeline.
In the description of the present application, it should be noted that, for the azimuth words such as "center", "lateral", "longitudinal", "length", "width", "thickness", "upper", "lower", "front", "rear", "left", "right", "vertical", "horizontal", "top", "bottom", "inner", "outer", "clockwise", "counterclockwise", etc., the azimuth and positional relationships are based on the azimuth or positional relationships shown in the drawings, it is merely for convenience of describing the present application and simplifying the description, and it is not to be construed as limiting the specific scope of protection of the present application that the device or element referred to must have a specific azimuth configuration and operation.
It should be noted that the terms "comprises" and "comprising," along with any variations thereof, in the description and claims of the present application are intended to cover a non-exclusive inclusion, such as a process, method, system, article, or apparatus that comprises a list of steps or elements is not necessarily limited to those steps or elements expressly listed or inherent to such process, method, article, or apparatus.
Note that the above is only a preferred embodiment of the present application and uses technical principles. It will be understood by those skilled in the art that the present application is not limited to the particular embodiments described herein, but is capable of various obvious changes, rearrangements and substitutions as will now become apparent to those skilled in the art without departing from the scope of the application. Therefore, while the present application has been described in connection with the above embodiments, it is to be understood that the application is not limited to the specific embodiments disclosed and that many other and equally effective embodiments may be devised without departing from the spirit of the application, and the scope thereof is determined by the scope of the appended claims.
Claims (10)
1. The buried pipeline perimeter collapse early warning method based on the water hammer characteristic parameter set is characterized by comprising the following steps of: the method comprises the following steps:
s1, uniformly defining and classifying defects by utilizing different peripheral collapse degrees of the buried pipeline, and dividing the early collapse of the buried pipeline into three conditions of bottom collapse, exposed surface and complete suspension; then collecting vibration data of the experiment platform pipeline system by using an acceleration sensor; collecting axial vibration data of a water hammer, wherein the axial direction is the extending direction of a pipeline;
s2, carrying out data batch preprocessing on vibration data collected by a pipeline system to obtain sample data, wherein denoising and data segmentation are adopted in the preprocessing;
s3, extracting features: extracting time domain sensitive characteristic information and frequency domain characteristics of water hammer vibration signals caused by early collapse of different buried pipelines in the preprocessed sample data to form a characteristic set; the time domain sensitive characteristic information is obtained by establishing a finite element simulation model by COMSOL, simulating valve closing action by setting a valve closing function to enable a pipeline to generate water hammer vibration waves, and carrying out pickup analysis on the water hammer vibration wave information by a point probe;
s4, according to the result of the COMSOL simulation model, closing the electromagnetic valve mainly to generate a water hammer and enabling the electromagnetic valve metal ball to hit the pipeline to generate vibration waves, and extracting vibration signal characteristics which are sensitive to the early collapse type of the perimeter of the flow transmission pipeline in the vicinity of 450 HZ;
s5, constructing different machine learning basic models, and training a support vector machine, a random forest, a radial basis neural network and a BP neural network; cross verification is adopted to verify the accuracy of the model, so that a machine learning model which is most suitable for further optimization is selected;
and S6, further optimizing the radial basis function neural network model, namely optimizing the parameter setting of the radial basis function neural network mainly through an ant colony algorithm.
S7, training the optimized machine learning model to obtain a parameter set with the highest recognition rate;
and S8, testing the parameter set with the highest recognition rate obtained in the step S7, and respectively recognizing early collapse to obtain the type of the early collapse of the perimeter of the flow transmission pipeline.
2. The method for pre-warning the perimeter collapse of the buried fluid delivery pipeline based on the water hammer characteristic parameter set according to claim 1, wherein the time domain characteristic parameters and expressions extracted in the step S3 are as follows:
wherein x (N) represents a time domain sequence of the signal, n=1, 2, …, N and N are the number of sample points, the signal is described by an average value A1, and the energy of the water hammer vibration signal under different conditions can be detected; the variance A2 represents the dynamic component of the signal energy, reflects the discrete degree between data, and has better accuracy in the aspects of model prediction and experimental data description; the effective value A3 describes the energy of the vibration signal;
the specific characteristic frequency domain parameter formula is as follows:
wherein S (K) represents a frequency domain sequence of the signal, where k=1, 2, …, K is the number of sample points; frequency domain feature statistics parameters are introduced to describe the characteristics of the signal: the spectral mean B1 describes its fluctuations, and the frequency distribution of the pipe vibration signal is described by the average frequency B2 and the root mean square frequency B3;representing the square of the corresponding frequency domain sequence.
3. The method for pre-warning the perimeter collapse of the buried fluid transport pipeline based on the characteristic parameter set of the water hammer wave according to claim 1, wherein the characteristic set in the step S3 comprises: bottom collapse, exposed surface and complete suspension, and time and frequency domain feature sets of the three conditions.
4. The method for pre-warning the perimeter collapse of the buried pipeline based on the water hammer characteristic parameter set according to claim 1, wherein in the step S4, when the characteristic of the sensitive vibration signal is extracted, the signal is decomposed and reconstructed by adopting a wave packet technology to obtain a desired wavelet packet frequency band, and then the frequency domain characteristic parameter is extracted; the specific wavelet packet selects three layers of wavelets, which are divided into 8 frequency bands, and Haar is a wavelet basis function.
5. The method for pre-warning the perimeter collapse of the buried pipeline based on the water hammer characteristic parameter set according to claim 1, wherein in the step S5, the MATLAB deep learning tool library is referenced for preliminary training test, and the most suitable machine learning method is selected for testing optimization, and the method comprises the following main steps:
data preparation: training data are arranged into a feature matrix X and a corresponding label vector Y; ensuring that the data has been standardized or preprocessed;
support vector machine training: creating a support vector machine model object using the fitcvm function; training a model by using a train method, and transmitting a feature matrix X and a label vector Y; the new data can be classified and predicted by using a prediction method;
random forest training: creating a random forest object by using the TreeBagger function; training a random forest model by using a train method, and transmitting a feature matrix X and a label vector Y; the new data can be classified and predicted by using a prediction method;
radial basis function neural network training: creating a radial basis function neural network object using the newgrnn function; training a network model by using a train method, and transmitting a feature matrix X and a label vector Y; the sim method can be used for carrying out classified prediction on new data;
BP neural network training: creating a BP neural network object by using a feedback forwarding net function; setting a network structure and training parameters; training a network model by using a train method, and transmitting a feature matrix X and a label vector Y; classification predictions can be made for new data using sim methods.
6. The method for pre-warning the perimeter collapse of the buried pipeline based on the water hammer characteristic parameter set according to claim 1, wherein in the step S6, the radial basis function neural network structure mainly comprises an input layer, a radial base layer and an output layer; the network structure relation is mainly adjusted according to the training set, and then the center point and the width parameters are determined.
7. The method for pre-warning the perimeter collapse of the buried fluid delivery pipeline based on the water hammer characteristic parameter set according to claim 6, wherein the step S6 specifically comprises the following steps:
s61: collecting and collating pipeline data for training and testing,
s62: preprocessing data, and extracting various characteristic values;
s63: establishing a basic radial basis function neural network, including initialization of the structure, weight and bias of the network and radial basis functions, S64: iterative optimization is carried out on hidden layers and super parameters of the radial base network through an ant colony algorithm, and S65: determining a network structure;
s66: further optimizing parameters by a backprojection algorithm.
8. The method for pre-warning the perimeter collapse of the buried fluid delivery pipeline based on the water hammer characteristic parameter set, which is disclosed by claim 7, is characterized in that: the step S64 specifically includes the following steps:
initializing parameters of an ant colony algorithm, including the number of ants, the volatilization rate of pheromones and the initial concentration of the pheromones; setting a proper numerical value according to actual conditions to perform ant iterative optimization;
the specific process is as follows:
s64.1 initializing ant colony parameters: setting the number and initial positions of ants, determining the scale of an ant colony, and distributing an initial position for each ant randomly or according to a specific strategy; the weight and bias of each ant are randomly initialized and used as the initial parameters of the neural network;
s64.2 initializing a pheromone matrix: creating an pheromone matrix corresponding to the problem scale, wherein the initial value can be set to be the same constant or can be adjusted according to the characteristics of the problem;
s64.3 routing of ants: each ant carries out forward propagation on a training data set according to the current neural network parameters, calculates the value of a loss function, and updates the neural network parameters according to the pheromone concentration and the heuristic function;
s64.4 pheromone update: updating the pheromone concentration matrix according to the path quality of each ant; a better path will increase the pheromone concentration and a worse path will decrease the pheromone concentration;
s64.5, repeatedly performing iterative optimization, and judging termination conditions: setting iteration times or stopping conditions: setting proper iteration times or stopping conditions according to practical problems and experience, for example, reaching the maximum iteration times and meeting convergence conditions; judging whether a termination condition is met;
s64.6 updating the neural network parameters: in the iteration process, recording and storing the optimal solution in each iteration, and updating the weight and bias of the neural network according to the neural network parameters corresponding to the optimal path;
s64.7, returning to the neural network model subjected to the ant colony algorithm depth optimization.
9. A computer-readable storage medium having a computer program stored thereon, characterized by: when the computer program is executed by a processor, the equipment where the computer readable storage medium is located executes the buried pipeline perimeter collapse early warning method based on the water hammer characteristic parameter set according to any one of claims 1 to 8.
10. An electronic device, characterized in that: the method comprises a processor and a memory, wherein the memory stores a program which can run on the processor, and the processor realizes the buried pipeline perimeter collapse early warning method based on the machine learning of the water hammer characteristic parameter set according to any one of 1-8 when executing the program.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311079873.1A CN117009788A (en) | 2023-08-25 | 2023-08-25 | Buried fluid delivery pipeline perimeter collapse early warning method, storage medium and method based on water hammer characteristic parameter set |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311079873.1A CN117009788A (en) | 2023-08-25 | 2023-08-25 | Buried fluid delivery pipeline perimeter collapse early warning method, storage medium and method based on water hammer characteristic parameter set |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN117009788A true CN117009788A (en) | 2023-11-07 |
Family
ID=88569129
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311079873.1A Pending CN117009788A (en) | 2023-08-25 | 2023-08-25 | Buried fluid delivery pipeline perimeter collapse early warning method, storage medium and method based on water hammer characteristic parameter set |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117009788A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117354042A (en) * | 2023-11-14 | 2024-01-05 | 龙坤(无锡)智慧科技有限公司 | Method for monitoring abnormal flow of edge gateway equipment in dynamic monitoring mode |
-
2023
- 2023-08-25 CN CN202311079873.1A patent/CN117009788A/en active Pending
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117354042A (en) * | 2023-11-14 | 2024-01-05 | 龙坤(无锡)智慧科技有限公司 | Method for monitoring abnormal flow of edge gateway equipment in dynamic monitoring mode |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108932480B (en) | Distributed optical fiber sensing signal feature learning and classifying method based on 1D-CNN | |
| Yang et al. | Novel leakage detection by ensemble 1DCNN-VAPSO-SVM in oil and gas pipeline systems | |
| Lei et al. | A comprehensive evaluation method for indoor air quality of buildings based on rough sets and a wavelet neural network | |
| Hung et al. | An artificial neural network model for rainfall forecasting in Bangkok, Thailand | |
| Curilem et al. | Classification of seismic signals at Villarrica volcano (Chile) using neural networks and genetic algorithms | |
| Wang et al. | A novel structural damage identification scheme based on deep learning framework | |
| CN110242865B (en) | Gas leakage detection and judgment method and system easy for continuous optimization | |
| Jalalkamali | Using of hybrid fuzzy models to predict spatiotemporal groundwater quality parameters | |
| Yang et al. | Hybrid two‐step method of damage detection for plate‐like structures | |
| CN117009788A (en) | Buried fluid delivery pipeline perimeter collapse early warning method, storage medium and method based on water hammer characteristic parameter set | |
| CN108399434B (en) | Analysis and prediction method of high-dimensional time series data based on feature extraction | |
| CN111520615A (en) | Pipe network leakage identification and positioning method based on line spectrum pair and cubic interpolation search | |
| CN104316167B (en) | Structural response method for detecting abnormality under a kind of low signal-to-noise ratio | |
| Chang et al. | Explore training self-organizing map methods for clustering high-dimensional flood inundation maps | |
| Xiong et al. | A novel deep convolutional image-denoiser network for structural vibration signal denoising | |
| Xie et al. | Autoencoder-based deep belief regression network for air particulate matter concentration forecasting | |
| Ji et al. | Structural performance degradation identification of offshore wind turbines based on variational mode decomposition with a Grey Wolf Optimizer algorithm | |
| Mostafaei et al. | Fully automated operational modal identification of regular and irregular buildings with ensemble learning | |
| CN113780094B (en) | Perimeter intrusion identification method based on Gaussian mixture model and hidden Markov model | |
| CN114818809A (en) | Fault diagnosis method for SCA-SVM motor rolling bearing based on cross wavelets | |
| Zhang et al. | Structural Damage Identification System Suitable for Old Arch Bridge in Rural Regions: Random Forest Approach. | |
| Khairudin et al. | In-Depth review on machine learning models for long-term flood forecasting | |
| CN117233347B (en) | Carbon steel spheroidization grade measuring method, system and equipment | |
| Zhan et al. | A Novel Structural Damage Detection Method via Multisensor Spatial–Temporal Graph-Based Features and Deep Graph Convolutional Network | |
| Tu et al. | A coupling model of multi-feature fusion and multi-machine learning model integration for defect recognition |
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 |