CN117668662A - Pipe network data processing method and system based on deep neural network model - Google Patents
Pipe network data processing method and system based on deep neural network model Download PDFInfo
- Publication number
- CN117668662A CN117668662A CN202311686371.5A CN202311686371A CN117668662A CN 117668662 A CN117668662 A CN 117668662A CN 202311686371 A CN202311686371 A CN 202311686371A CN 117668662 A CN117668662 A CN 117668662A
- Authority
- CN
- China
- Prior art keywords
- model
- monitoring data
- data
- monitoring
- pipe network
- 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
- 238000003062 neural network model Methods 0.000 title claims abstract description 35
- 238000003672 processing method Methods 0.000 title claims abstract description 14
- 238000012544 monitoring process Methods 0.000 claims abstract description 129
- 238000000034 method Methods 0.000 claims abstract description 34
- 238000012549 training Methods 0.000 claims abstract description 31
- 230000002159 abnormal effect Effects 0.000 claims abstract description 26
- 238000012545 processing Methods 0.000 claims abstract description 17
- 238000004458 analytical method Methods 0.000 claims abstract description 12
- 238000012417 linear regression Methods 0.000 claims description 21
- 238000012795 verification Methods 0.000 claims description 12
- 238000013473 artificial intelligence Methods 0.000 claims description 10
- VNWKTOKETHGBQD-UHFFFAOYSA-N methane Chemical compound C VNWKTOKETHGBQD-UHFFFAOYSA-N 0.000 claims description 8
- 239000003345 natural gas Substances 0.000 claims description 4
- 238000012067 mathematical method Methods 0.000 abstract description 4
- 238000012937 correction Methods 0.000 abstract description 2
- 238000004590 computer program Methods 0.000 description 8
- 239000007789 gas Substances 0.000 description 6
- 230000008569 process Effects 0.000 description 6
- 238000010586 diagram Methods 0.000 description 5
- 238000010276 construction Methods 0.000 description 4
- 238000005516 engineering process Methods 0.000 description 4
- 238000012216 screening Methods 0.000 description 4
- 238000004088 simulation Methods 0.000 description 4
- 230000009471 action Effects 0.000 description 3
- 238000012806 monitoring device Methods 0.000 description 3
- 238000004891 communication Methods 0.000 description 2
- 238000013499 data model Methods 0.000 description 2
- 238000011161 development Methods 0.000 description 2
- 230000018109 developmental process Effects 0.000 description 2
- 238000013515 script Methods 0.000 description 2
- 230000005856 abnormality Effects 0.000 description 1
- 230000032683 aging Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 238000007405 data analysis Methods 0.000 description 1
- 230000009977 dual effect Effects 0.000 description 1
- 238000004880 explosion Methods 0.000 description 1
- 230000006870 function Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 230000000750 progressive effect Effects 0.000 description 1
- 230000008439 repair process Effects 0.000 description 1
- 238000010200 validation analysis Methods 0.000 description 1
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 description 1
Landscapes
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention relates to a pipe network data processing method and system based on a deep neural network model, wherein the method comprises the following steps: acquiring the type of a pipe network node and historical monitoring data thereof; constructing a sample set and training an intelligent positioning model; the intelligent positioning model is a deep neural network model; continuously collecting real-time monitoring data at a plurality of nearest time points; and inputting the monitoring data into an intelligent positioning model corresponding to the type of the monitoring data to obtain an intelligent positioning result. According to the invention, the traditional mathematical method is utilized to obtain the available training samples, the specific pipe network structure is served based on the neural network model, and the use threshold of the neural network model in the pipe network data processing field is reduced through the mutual correction of the normal sample and the abnormal sample, so that the analysis efficiency of pipe network data is greatly improved.
Description
[ field of technology ]
The invention belongs to the technical field of intelligent analysis and processing, and particularly relates to a pipe network data processing method and system based on a deep neural network model.
[ background Art ]
With the rapid development of urban construction, the systems of urban water supply network, gas pipe network and the like are gradually huge and more complex, and the pipe explosion event occurs at time, so that the daily life of people is seriously influenced. Furthermore, the range of gas supply is too large, the pipeline is too long, and the distribution of gas supply points is relatively large. In addition, the gas pipeline is mostly buried underground, the joints and valves of the pipeline are more, and the gas leakage phenomenon caused by gas leakage in the conveying process is unavoidable. If equipment accidents are caused by equipment aging and undetected in place, and then accidents such as leakage of a large amount of natural gas are caused, the safety of human bodies and social property are threatened greatly, and therefore emergency repair is very important in time.
On the one hand, artificial intelligence is a branch of computer science, and english is abbreviated as AI, which is a new technical science for researching and developing theories, methods, technologies and application systems for simulating, extending and expanding human intelligence. With the development of technology, particularly in recent years, artificial intelligence technology has been fully applied to various technical fields to improve the leading performance of the technical fields;
in another aspect, a method for identifying pipe network anomalies includes: the traditional mathematical method combines the subjects such as probability theory, statistics and the like to carry out equation fitting on data, a simulation model and an emerging artificial intelligent model; however, the accuracy of conventional mathematical methods is actually too low and is now rarely used. Simulation models are currently the most popular, with some accuracy. However, the building of the simulation model is very complicated, and a long process is required to obtain a qualified simulation model, so that the cost is high. The accuracy of artificial intelligence models comes largely from the training and validation process, which requires a large number of actual samples, especially for deep neural network models with very high accuracy. According to the invention, the traditional mathematical method is utilized to obtain the available training samples, the specific pipe network structure is served based on the neural network model, and the use threshold of the neural network model in the pipe network data processing field is reduced through the mutual correction of the normal sample and the abnormal sample, so that the analysis efficiency of pipe network data is greatly improved.
[ invention ]
In order to solve the above problems in the prior art, the present invention provides a pipe network data processing method and system based on a deep neural network model, where the method includes:
step S1: acquiring the type of a pipe network node and historical monitoring data thereof; sorting the monitoring data based on the monitoring data types, grouping the monitoring data according to a time sequence, wherein one time point corresponds to one group of monitoring data; each group of monitoring data of a type comprises monitoring data of all monitoring nodes numbered according to a preset sequence;
step S2: aiming at various types of monitoring data, respectively constructing and solving a linear regression model based on the monitoring data; the method comprises the following steps: for each monitored data type, a multiple linear regression model y=βx+epsi1=β is solved for the monitored data 0 +β 1 x 1 +β 2 x 2 +…+β k x k +…+β K x K +ε1; wherein: y is node No. 0 monitoring data, β= (β) 0 ,…β k ,…β K ) Is a regression parameter, x= (X) 1 ,…x k ,…x K ) Is the monitoring data of nodes 1 to K, wherein: k is 1 to K; ε 1 is the estimated error; when node No. 0 is 1, y= (Y) 0 ) The method comprises the steps of carrying out a first treatment on the surface of the The number of pipe network nodes is K+1;
step S3: aiming at various types of monitoring data, respectively creating and solving a propagation structure model based on the monitoring data difference value; setting a propagation structure model: y=α 1 Δx 1 +(α 1 ,α 2 )(Δx 1 ,Δx 2 )…+(α 1 ,…,α k )(Δx 1 ,…,Δx k )…+(α 1 ,…,α K )(Δx 1 ,…,Δx K ) +ε2; wherein: (Deltax) 1 ,…Δx k ,…Δx K ) Is the difference value of the monitoring data of the No. 1-K node and the No. 0 node, wherein: k is 1 to K; when node No. 0 is 1, y= (Y) 0 ) Node 0 monitoring data; (alpha) 0 ,…α k ,…α K ) Respectively the propagation coefficients; ε 2 is the propagation reference;
step S4: constructing a training data set containing an abnormal sample and a normal sample for training the intelligent positioning model based on historical and real-time monitoring data, a linear regression model and a propagation structure model; training the intelligent positioning model by using the training data set until a training target is reached; constructing a verification data set for verifying the intelligent positioning model based on the historical monitoring data; when the verification result of the intelligent positioning model meets the verification target, entering the next step; otherwise, repeating the step; repeating the step S4 aiming at different types of monitoring data to obtain intelligent positioning models which correspond to different monitoring data types and accord with training and verification targets;
step S5: continuously collecting real-time monitoring data at a plurality of nearest time points; and inputting the monitoring data into an intelligent positioning model corresponding to the type of the monitoring data to obtain an intelligent positioning result.
Further, the pipe network is a natural gas pipe network.
Further, the monitoring device for monitoring the node is one or more of a pressure, a flow, and a sound monitoring device.
Further, a linear regression model is solved by a least square method.
Further, the smart positioning model is an artificial smart model.
Further, the artificial intelligence model is a deep neural network model.
Further, the artificial intelligence model is a post-feedback neural network model.
Further, the smart positioning model is a ReLU model.
Further, the anomaly is a leak and/or a blockage.
A pipe network data processing system based on a deep neural network model, the system comprising: the monitoring terminal and the monitoring analysis server; the monitoring terminal is used for sending a pipe network data processing request to the monitoring analysis server; the monitoring analysis server is used for executing the pipe network data processing method based on the deep neural network model and sending the obtained positioning result to the monitoring terminal.
The beneficial effects of the invention include:
(1) The method has the advantages that a monitoring data model and a difference model which can be mutually corrected are arranged, the construction mode is simple, the calculated amount is small, a large number of samples can be quickly generated based on historical monitoring data, and a complex SCADA model based on a topological structure is not required to be constructed; meanwhile, the generation of a normal sample and an abnormal sample can be controlled in the direction according to the result of model training, and the training efficiency of the model is improved;
further: the structure information of the pipe network is overlapped on the numerical information by a sequential coding mode based on the No. 0 node, so that the propagation structure model can analyze and obtain more useful information, and meanwhile, the data formats of the linear regression model, the propagation structure model and the intelligent positioning model are unified on the basis of effectively utilizing double information by the numbering mode or the coding mode;
(2) Because sample data and model training are relatively easy, the process can be simply and repeatedly applied to various types of monitoring devices, the multi-dimensional positioning result based on the multi-monitoring type is constructed accurately, and the stability of the abnormal positioning result is improved through mixing and weighted mixing of the intelligent positioning result.
[ description of the drawings ]
The accompanying drawings, which are included to provide a further understanding of the invention and are incorporated in and constitute a part of this application, illustrate and together with the description serve to explain the invention, if necessary:
fig. 1 is a schematic diagram of a pipe network data processing method based on a deep neural network model.
[ detailed description ] of the invention
The present invention will now be described in detail with reference to the drawings and the specific embodiments thereof, wherein the exemplary embodiments and the description are for the purpose of illustrating the invention only and are not to be construed as limiting the invention.
As shown in fig. 1, the invention provides a pipe network data processing method based on a deep neural network model, which comprises the following steps:
step S1: acquiring the type of a pipe network node and historical monitoring data thereof; sorting the monitoring data based on the monitoring data types, grouping the monitoring data according to a time sequence, wherein one time point corresponds to one group of monitoring data; each set of monitoring data of a type includes monitoring data of all monitoring nodes numbered in sequence; setting a No. 0 node, sorting the nodes in the pipe network structure according to the distance from the No. 0 node, and numbering the nodes successively; that is, the closer to node 0, the smaller the number value, and conversely, the larger the number value; wherein: the monitoring data are historical and real-time data acquired through monitoring equipment arranged on pipe network nodes; for example: flow value (I) 0 ,I 1 ,I 2 ) A monitoring data set for a traffic type for a pipe network structure comprising 3 nodes, where k=2; the nodes with topological connection relations in the pipe network structure monitor the nodes in multiple types, so the nodes are also called monitoring nodes;
preferably: the distance is calculated by physical distance, topological distance, distance by the number of nodes and transmission loss (the larger the loss is, the farther the distance is, and conversely the smaller the loss is, the closer the distance is);
preferably: the pipe network is a natural gas pipe network; the monitoring equipment is pressure, flow, sound monitoring equipment, image acquisition equipment and the like;
step S2: aiming at various types of monitoring data, respectively constructing and solving a linear regression model based on the monitoring data; the method comprises the following steps: for each monitored data type, a multiple linear regression model y=βx+epsi1=β is solved for the monitored data 0 +β 1 x 1 +β 2 x 2 +…+β k x k +…+β K x K +ε1; wherein: y is node No. 0 monitoring data, β= (β) 0 ,…β k ,…β K ) Is a regression parameter, x= (X) 1 ,…x k ,…x K ) Is the monitoring data of nodes 1 to K, wherein: k is 1 to K; ε 1 is the estimated error; when node No. 0 is 1, y= (Y) 0 ) The method comprises the steps of carrying out a first treatment on the surface of the When node 0 is N 0 When y= (Y) 0,1 …y 0,N0 )；That is, the number of pipe network nodes is K+1;
setting a No. 0 node according to the type of the node, the size of the monitoring data, the position in the pipe network structure and the like; for example: setting a No. 0 node as a heat source node;
preferably: solving a linear regression model by using a least square method;
further: the structure information of the pipe network is overlapped on the numerical information by a sequential coding mode based on the No. 0 node, so that the propagation structure model can analyze and obtain more useful information, and meanwhile, the data formats of the linear regression model, the propagation structure model and the intelligent positioning model are unified on the basis of effectively utilizing double information by the numbering mode or the coding mode;
step S3: aiming at various types of monitoring data, respectively creating and solving a propagation structure model based on the monitoring data difference value; setting a propagation structure model: y=α 1 Δx 1 +(α 1 ,α 2 )(Δx 1 ,Δx 2 )…+(α 1 ,…,α k )(Δx 1 ,…,Δx k )…+(α 1 ,…,α K )(Δx 1 ,…,Δx K ) +ε2; wherein: (Deltax) 1 ,…Δx k ,…Δx K ) Is the difference value of the monitoring data of the No. 1-K node and the No. 0 node, wherein: k is 1 to K; y is node number 0 monitoring data; (alpha) 0 ,…α k ,…α K ) Respectively the propagation coefficients; ε 2 is the propagation reference; the variable of the propagation structure model can find more structural characteristics and better characterize the difference variable;
preferably: solving a propagation structure model through monitoring data fitting and a polynomial equation;
preferably: epsilon 1 and epsilon 2 are set to preset values;
step S4: constructing a training data set containing an abnormal sample and a normal sample for training the intelligent positioning model based on historical and real-time monitoring data, a linear regression model and a propagation structure model; training the intelligent positioning model by using the training data set until a training target is reached; constructing a verification data set for verifying the intelligent positioning model based on the historical monitoring data; when the verification result of the intelligent positioning model meets the verification target, entering the next step; otherwise, repeating the step; repeating the step S4 for different types of monitoring data to obtain intelligent positioning models corresponding to different monitoring data types; the intelligent positioning model is used for performing abnormal positioning;
preferably: the intelligent positioning model is an artificial intelligent model; for example: the artificial intelligence model is a deep neural network model;
alternatively, the following is used: the artificial intelligence model is a post-feedback neural network model;
preferably: the abnormality is leakage and blockage positioning;
preferably: when the training target cannot be reached after training or the verification target cannot be reached after verification, distinguishing a normal sample from an abnormal sample, when the target cannot be reached for the normal sample, increasing the number of the normal samples in the training data set, and when the target cannot be reached for the abnormal sample, increasing the number of the abnormal samples in the training data set;
preferably: the method for constructing the training data set for training the intelligent positioning model, which comprises an abnormal sample and a normal sample, based on the historical monitoring data, the linear regression model and the propagation structure model comprises the following steps:
step S4A1: aiming at the construction of a normal sample, constructing a normal sample to be selected by using historical monitoring data and a linear regression model; screening a normal sample set to be selected by using the propagation structure model so as to retain normal samples; specific: on the basis of the historical monitoring data, setting a preset interpolation range for each element in each group of monitoring data in the historical monitoring data, so that data interpolation can be performed in the element values and the interpolation ranges thereof; wherein: the preset interpolation range is related to the element value proportion of each element; numbering the interpolation element values in sequence to form a normal sample to be selected; inputting a normal sample to be selected into a propagation structure model to judge whether the element value of the normal sample deviates from the propagation structure model, if so, screening the normal sample to be selected, otherwise, taking the normal sample to be selected as a reserved normal sample;
preferably: the preset interpolation range is within the range of 0% -5% of the element value;
step S4A2: constructing an abnormal sample, and constructing a normal sample set to be selected by using historical monitoring data and a propagation structure model; screening the to-be-selected abnormal samples by using a linear regression model to reserve the abnormal samples; specific: setting an interpolation range for each element in each group of monitoring data in the historical monitoring data on the basis of the historical monitoring data; setting the interpolation range of the nodes in a differentiated mode, wherein the interpolation range of the locating points of the nodes at the abnormal locating positions is larger than the adjacent interpolation range of the adjacent nodes of the nodes, and the interpolation ranges of the locating points of the nodes at the abnormal locating positions are larger than the preset difference ranges of other nodes; the element value can conduct data interpolation in the element value and the interpolation range corresponding to the element value; wherein: the preset interpolation range is related to the element value proportion of the node element; numbering the interpolation element values in sequence to form an abnormal sample; inputting the to-be-selected abnormal sample into a linear regression model to judge whether the element value of the to-be-selected abnormal sample deviates from the linear regression model, if so, screening the to-be-selected abnormal sample, otherwise, taking the to-be-selected abnormal sample as a reserved abnormal sample;
preferably: the locating point interpolation range is within the range of 5% -10% of the element value; the adjacent interpolation range is 2% -5% of the element value; the interpolation range is within the range of 0% -2% of the element value;
step S4A3: judging whether the ratio of the number of the normal samples to the number of the abnormal samples reaches a preset ratio, judging whether the sum of the number of the normal samples and the number of the abnormal samples reaches the number requirement of the sample set, if so, ending, otherwise, returning to the step S4A1 or S4A2;
preferably: the preset proportion is 4-10:1, a step of;
according to the invention, the monitoring data model and the difference model which can be mutually corrected are formed, the construction mode is simple, the calculated amount is small, a large number of samples can be quickly generated based on historical monitoring data, and a complex SCADA model based on a topological structure is not required to be constructed; meanwhile, the generation of a normal sample and an abnormal sample can be controlled in the direction according to the result of model training, and the training efficiency of the model is improved; it can also be seen that the structure information of the pipe network is superimposed on the numerical information itself based on the sequential coding mode of the node 0, so that the propagation structure model can analyze and obtain more useful information, and meanwhile, the data formats of the linear regression model, the propagation structure model and the intelligent positioning model are unified on the basis of effectively utilizing the dual information through the numbering mode or the coding mode;
step S5: continuously collecting real-time monitoring data at a plurality of nearest time points; sorting the real-time monitoring data based on the monitoring data type; respectively inputting each type of monitoring data into an intelligent positioning model corresponding to the type of the monitoring data to obtain an intelligent positioning result set corresponding to one type of the monitoring data; obtaining a final intelligent positioning result based on one or more intelligent positioning result sets;
the step S5 specifically includes the following steps:
step S51: continuously collecting real-time monitoring data at a plurality of nearest time points; sorting the real-time monitoring data based on the monitoring data type and time;
step S52: for one monitoring data type, the monitoring data belonging to the same time point are numbered according to the sequence to form the input data of the intelligent positioning model; inputting the input data into an intelligent positioning model corresponding to the type of the input data to obtain a positioning result; repeating the step until all the monitoring data of the monitoring data type are processed; the obtained positioning result forms an intelligent positioning result set corresponding to the type of the monitoring data; wherein: the sequential numbering mode is the same as the previous mode, and the nodes are respectively ordered according to the distance between the nodes 0 and the nodes 0;
step S53: mixing various intelligent positioning result sets, clustering the mixed results to obtain a clustering center, and taking the positioning result corresponding to the clustering center as an intelligent positioning result;
preferably: the mixing is that elements in the set are directly combined to form a set;
alternatively, the following is used: performing weighted mixing on various types of intelligent positioning result sets; the method comprises the following steps: verifying various types of intelligent positioning models to determine the accuracy rate, and determining the weight value of the corresponding type according to the accuracy rate, so that the weight value is higher when the accuracy rate is higher, and conversely, the weight value is lower when the accuracy rate is lower; in the clustering, the clustering center is more biased to the higher weight value by considering the weight value;
based on the same inventive concept, the invention provides a pipe network data processing system based on a deep neural network model, which comprises a monitoring terminal and a monitoring analysis server; the user terminal is in communication connection with the monitoring analysis server;
the monitoring analysis server is used for constructing and storing a linear regression model, a propagation structure model and an intelligent positioning model; the method is also used for executing the pipe network data analysis and processing method based on the artificial intelligence model;
preferably: the monitoring terminals are multiple; the monitoring terminal can be arranged on site; setting the mobile terminal as a monitoring terminal in a mode of setting a monitoring program;
based on the same inventive concept, the invention also provides a pipe network data processing electronic device based on the deep neural network model, which comprises the pipe network data processing method based on the deep neural network model; preferably, the electronic device includes a computer or the like;
in this document, relational terms such as first and second, and the like may be used solely to distinguish one entity or action from another entity or action without necessarily requiring or implying any actual such relationship or order between such entities or actions. Moreover, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising one … …" does not exclude the presence of other like elements in a process, method, article, or apparatus that comprises the element.
A computer program (also known as a program, software application, script, or code) can be written in any form of programming language, including compiled or interpreted languages, declarative or procedural languages, and it can be deployed in any form, including as a stand-alone program or as a module, component, subroutine, object or other unit suitable for use in a computing environment. The computer program may, but need not, correspond to a file in a file system. A program can be stored in a portion of a file that holds other programs or data (e.g., one or more scripts stored in a markup language document), in a single file dedicated to the program, or in multiple coordinated files (e.g., files that store one or more modules, subroutines, or portions of code). A computer program can be deployed to be executed on one computer or on multiple computers at one site or distributed across multiple sites and interconnected by a communication network.
It will be appreciated by those skilled in the art that embodiments of the present invention may be provided as a method, system, or computer program product. Accordingly, the present invention may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, the present invention may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, and the like) having computer-usable program code embodied therein.
The present invention is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems) and computer program products according to embodiments of the application. It will be understood that each flow and/or block of the flowchart illustrations and/or block diagrams, and combinations of flows and/or blocks in the flowchart illustrations and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
In the present specification, each embodiment is described in a progressive manner, and each embodiment is mainly described in a different point from other embodiments, and identical and similar parts between the embodiments are all enough to refer to each other.
Finally, it should be noted that: the above embodiments are only for illustrating the technical aspects of the present invention and not for limiting the same, and although the present invention has been described in detail with reference to the above embodiments, it should be understood by those of ordinary skill in the art that: modifications and equivalents may be made to the specific embodiments of the invention without departing from the spirit and scope of the invention, which is intended to be covered by the claims.
Claims (10)
1. A pipe network data processing method based on a deep neural network model is characterized by comprising the following steps:
step S1: acquiring the type of a pipe network node and historical monitoring data thereof; sorting the monitoring data based on the monitoring data types, grouping the monitoring data according to a time sequence, wherein one time point corresponds to one group of monitoring data; each group of monitoring data of a type comprises monitoring data of all monitoring nodes numbered according to a preset sequence;
step S2: aiming at various types of monitoring data, respectively constructing and solving a linear regression model based on the monitoring data; the method comprises the following steps: for each monitored data type, a multiple linear regression model y=βx+epsi1=β is solved for the monitored data 0 +β 1 x 1 +β 2 x 2 +…+β k x k +…+β K x K +ε1; wherein: y is node No. 0 monitoring data, β= (β) 0 ,…β k ,…β K ) Is a regression parameter, x= (X) 1 ,…x k ,…x K ) Is the monitoring data of nodes 1 to K, wherein: k is E1K is about; ε 1 is the estimated error; when node No. 0 is 1, y= (Y) 0 ) The method comprises the steps of carrying out a first treatment on the surface of the When node 0 is N 0 When y= (Y) 0,1 …y 0,N0 ) The method comprises the steps of carrying out a first treatment on the surface of the The number of pipe network nodes is K+1;
step S3: aiming at various types of monitoring data, respectively creating and solving a propagation structure model based on the monitoring data difference value; setting a propagation structure model: y=α 1 Δx 1 +(α 1 ,α 2 )(Δx 1 ,Δx 2 )…+(α 1 ,…,α k )(Δx 1 ,…,Δx k )…+(α 1 ,…,α K )(Δx 1 ,…,Δx K ) +ε2; wherein: (Deltax) 1 ,…Δx k ,…Δx K ) Is the difference value of the monitoring data of the No. 1-K node and the No. 0 node, wherein: k is 1 to K; when node No. 0 is 1, y= (Y) 0 ) Node 0 monitoring data; (alpha) 0 ,…α k ,…α K ) Respectively the propagation coefficients; ε 2 is the propagation reference;
step S4: constructing a training data set containing an abnormal sample and a normal sample for training the intelligent positioning model based on historical and real-time monitoring data, a linear regression model and a propagation structure model; training the intelligent positioning model by using the training data set until a training target is reached; constructing a verification data set for verifying the intelligent positioning model based on the historical monitoring data; when the verification result of the intelligent positioning model meets the verification target, entering the next step; otherwise, repeating the step; repeating the step S4 for different types of monitoring data to obtain intelligent positioning models corresponding to different monitoring data types;
step S5: continuously collecting real-time monitoring data at a plurality of nearest time points; and inputting the monitoring data into an intelligent positioning model corresponding to the type of the monitoring data to obtain an intelligent positioning result.
2. The pipe network data processing method based on the deep neural network model of claim 1, wherein the pipe network is a natural gas pipe network.
3. The method for processing pipe network data based on the deep neural network model according to claim 2, wherein the monitoring equipment used for monitoring the nodes is one or more of pressure, flow and sound monitoring equipment.
4. A pipe network data processing method based on a deep neural network model according to claim 3, wherein the linear regression model is solved by a least square method.
5. The method for processing pipe network data based on the deep neural network model according to claim 4, wherein the intelligent positioning model is an artificial intelligent model.
6. The method for processing pipe network data based on the deep neural network model according to claim 5, wherein the artificial intelligence model is a deep neural network model.
7. The network data processing system based on the deep neural network model of claim 5, wherein the artificial intelligence model is a post-feedback neural network model.
8. The method for processing pipe network data based on a deep neural network model according to claim 4, wherein the intelligent positioning model is a ReLU model.
9. The method for processing pipe network data based on the deep neural network model according to claim 4, wherein the anomaly is leakage and/or blockage.
10. A pipe network data processing system based on a deep neural network model, the system comprising: the monitoring terminal and the monitoring analysis server; the monitoring terminal is used for sending a pipe network data processing request to the monitoring analysis server; the monitoring analysis server is used for executing the pipe network data processing method based on the deep neural network model according to any one of claims 1-9, and sending the obtained positioning result to the monitoring terminal.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311686371.5A CN117668662A (en) | 2023-12-08 | 2023-12-08 | Pipe network data processing method and system based on deep neural network model |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311686371.5A CN117668662A (en) | 2023-12-08 | 2023-12-08 | Pipe network data processing method and system based on deep neural network model |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN117668662A true CN117668662A (en) | 2024-03-08 |
Family
ID=90065892
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311686371.5A Pending CN117668662A (en) | 2023-12-08 | 2023-12-08 | Pipe network data processing method and system based on deep neural network model |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117668662A (en) |
-
2023
- 2023-12-08 CN CN202311686371.5A patent/CN117668662A/en active Pending
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109886830B (en) | Water supply network pollution source tracking and positioning method based on user complaint information | |
| Palau et al. | Burst detection in water networks using principal component analysis | |
| CN112348292B (en) | Short-term wind power prediction method and system based on deep learning network | |
| CN110909485B (en) | SWMM model parameter self-calibration method based on BP neural network | |
| CN102072409B (en) | Pipe network leakage monitoring method combining leakage probability calculation and recorder monitoring | |
| CN112799898B (en) | Interconnection system fault node positioning method and system based on distributed fault detection | |
| Geng et al. | A fault detection method based on horizontal visibility graph‐integrated complex networks: Application to complex chemical processes | |
| CN116862081B (en) | Operation and maintenance method and system for pollution treatment equipment | |
| Li et al. | Fast detection and localization of multiple leaks in water distribution network jointly driven by simulation and machine learning | |
| Yang et al. | Resilience assessment and improvement for electric power transmission systems against typhoon disasters: a data-model hybrid driven approach | |
| CN114548680A (en) | Method and system for automatically calibrating parameters of urban storm flood management model | |
| CN114198644A (en) | DMA (direct memory access) monitoring-based water supply network leakage detection control method for related flow data | |
| CN112948757A (en) | Low-voltage distribution area topology verification method based on improved Pearson correlation coefficient | |
| CN110716998A (en) | Method for spatializing fine-scale population data | |
| CN113486950A (en) | Intelligent pipe network water leakage detection method and system | |
| CN117520989A (en) | Natural gas pipeline leakage detection method based on machine learning | |
| CN117668662A (en) | Pipe network data processing method and system based on deep neural network model | |
| CN112862063A (en) | Complex pipe network leakage positioning method based on deep belief network | |
| CN103646095B (en) | The reliability of a kind of common cause failure based on data-driven judges system and method | |
| CN113970073B (en) | ResNet-based water supply network leakage accurate positioning method | |
| CN113779892B (en) | Method for predicting wind speed and wind direction | |
| CN115493093A (en) | Steam heating pipe network leakage positioning method and system based on mechanical simulation | |
| CN103337000A (en) | Safety monitoring and prewarning method for oil-gas gathering and transferring system | |
| CN109408604B (en) | Grid processing method, device and system for data of weather factors associated with power transmission line | |
| Gupta et al. | Modeling and simulation of CEERI's water distribution network to detect leakage using HLR approach |
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 |