CN108804740B - Long-distance pipeline pressure monitoring method based on integrated improved ICA-KRR algorithm - Google Patents
Long-distance pipeline pressure monitoring method based on integrated improved ICA-KRR algorithm Download PDFInfo
- Publication number
- CN108804740B CN108804740B CN201810285077.6A CN201810285077A CN108804740B CN 108804740 B CN108804740 B CN 108804740B CN 201810285077 A CN201810285077 A CN 201810285077A CN 108804740 B CN108804740 B CN 108804740B
- Authority
- CN
- China
- Prior art keywords
- matrix
- long
- component
- pressure monitoring
- distance pipeline
- 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.)
- Active
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
-
- F—MECHANICAL ENGINEERING; LIGHTING; HEATING; WEAPONS; BLASTING
- F17—STORING OR DISTRIBUTING GASES OR LIQUIDS
- F17D—PIPE-LINE SYSTEMS; PIPE-LINES
- F17D5/00—Protection or supervision of installations
- F17D5/02—Preventing, monitoring, or locating loss
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F17/00—Digital computing or data processing equipment or methods, specially adapted for specific functions
- G06F17/10—Complex mathematical operations
- G06F17/15—Correlation function computation including computation of convolution operations
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F17/00—Digital computing or data processing equipment or methods, specially adapted for specific functions
- G06F17/10—Complex mathematical operations
- G06F17/16—Matrix or vector computation, e.g. matrix-matrix or matrix-vector multiplication, matrix factorization
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F17/00—Digital computing or data processing equipment or methods, specially adapted for specific functions
- G06F17/10—Complex mathematical operations
- G06F17/18—Complex mathematical operations for evaluating statistical data, e.g. average values, frequency distributions, probability functions, regression analysis
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2113/00—Details relating to the application field
- G06F2113/14—Pipes
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2119/00—Details relating to the type or aim of the analysis or the optimisation
- G06F2119/06—Power analysis or power optimisation
Abstract
The invention discloses a long-distance pipeline pressure monitoring method based on an integrated and improved ICA-KRR algorithm, which comprises the following steps: 1) Constructing a pressure monitoring data matrix of the long-distance pipeline; 2) Calculation variable P ε R m×m The method comprises the steps of carrying out a first treatment on the surface of the 3) Extracting component matrix t=p T X is a group; 4) Whitening the extracted component matrix T to obtain whitened results; 5) Calculating a matrix s=c T Z; 6) Calculate matrix C n The method comprises the steps of carrying out a first treatment on the surface of the 7) Calculating a separation matrix W E R d×m Mixing matrix A epsilon R m×d The method comprises the steps of carrying out a first treatment on the surface of the 8) Obtaining source signals of independent components, and the source signals of the independent components are connectedThe independent relation of (2) is reflected by non-gaussian property, the non-gaussian property is quantized by a negative entropy function, and the negative entropy function can select three non-quadratic functions; 9) Constructing three component importance evaluation criteria; 10 Mixing to form a double-layer comprehensive learning strategy; 11 To form 9 component selection models; 12 Obtaining a weight coefficient w;13 Obtaining regression fault signal data y;14 The leakage position d is calculated, and the method can realize real-time monitoring and accurate positioning of the leakage position on the long-distance pipeline.
Description
Technical Field
The invention belongs to the technical field of long oil and gas pipeline safety detection, and relates to a long-distance pipeline pressure monitoring method based on an integrated and improved ICA-KRR algorithm.
Background
In recent years, as the service life of oil and gas pipelines increases, the frequency of pipeline leakage accidents caused by various non-human factors is continuously increased, and serious economic losses are brought to enterprises. Therefore, the method has important research significance in monitoring pipeline pressure in real time, accurately positioning leakage faults and timely early warning.
The oil gas pipeline leakage detection technology is an important means for guaranteeing the safe production of pipelines. Along with the rapid development of information technology and modern control theory, the leakage detection method has been widely applied in a high-efficiency and flexible application mode, and has become a main means for continuously monitoring the leakage of the pipeline. Because of the strong burst of pipe leakage, pressure monitoring data is complex and redundant. Therefore, the pipeline pressure is monitored, leakage faults are timely early-warned, effective information and characteristic rules in pipeline pressure monitoring data are required to be identified, and a proper method and a proper model are established to enable a pressure data analysis party to realize fault detection.
In this regard, pressure monitoring and risk early warning work of long-distance oil and gas pipelines are moving toward quantitative and proactive strategies. At present, students at home and abroad have more researches on the aspect of pipeline leakage faults, and various characteristic research methods are provided for pipeline pressure monitoring and fault leakage positioning. Wang Mingda and the like utilize independent component analysis to combine the support vector machine to realize leakage detection on noise reduction and separation of pipeline pressure signals, zhang Yu and the like respectively adopt a dynamic pressure transmitter to measure dynamic change of pipeline pressure and an empirical mode decomposition method to analyze pressure signals, but data analysis on fault pressure signals cannot realize accurate positioning, lin Weiguo and the like utilize wavelet denoising to identify and extract abnormal signals, so that screening of interference signals and leakage signals is realized, the effectiveness of a pipeline multipoint pressure monitoring scheme is not high, lu Sheng and the like adopt independent component analysis and principal component analysis to separate signals and combine a Lasso regression method to realize positioning and selection of main abnormal variables when faults occur, abnormal variables which generate faults are detected step by step and realize positioning, but the Lasso regression is not ideal for the processing speed effect of a medium-sized data set, chudong is based on an independent component analysis model, double-layer Bayesian reference is expanded, and selection of independent components is strengthened. In summary, the existing research method has limitations of different degrees on pipeline pressure monitoring and leakage fault positioning, and the positioning accuracy and the actual condition are not ideal.
Based on the above analysis, the authors therefore propose a pipeline pressure monitoring and leakage real-time localization integrated model combining a comprehensive improved independent component analysis algorithm (Ensemble Modified Independent Component Analysis, EMICA) with a kernel-ridge regression algorithm (Kernel Ridge Regression, KRR). Training EMICA model with normal offline data to build statistics T 2 And Q constraint and improves model efficiency. And carrying out regression analysis on the data before and after the fault by the KRR algorithm to obtain the variation amplitude of the regression coefficient of the leakage sequence, thereby realizing the positioning and diagnosis of the leakage. Finally, a numerical simulation experiment of TE (Tennessee Eastman) procedure was performed and compared with existing leak diagnostic methods to verify the performance of the proposed method.
The method has a certain effect through different pressure monitoring methods, but the problems of poor discrimination capability and low accuracy caused by the fact that the original number sequence is not selected and the applicability analysis is insufficient in the application are also existed in the application.
Disclosure of Invention
The invention aims to overcome the defects of the prior art, and provides a long-distance pipeline pressure monitoring method based on an integrated and improved ICA-KRR algorithm, which can realize real-time monitoring and accurate positioning of the leakage position on a long-distance pipeline.
In order to achieve the above purpose, the method for monitoring the pressure of the long-distance pipeline based on the integrated and improved ICA-KRR algorithm comprises the following steps:
1) Collecting pressure monitoring data of a plurality of long-distance pipelines, and constructing a long-distance pipeline pressure monitoring data matrix X epsilon R by the obtained pressure monitoring data of the long-distance pipelines m×n N is the number of obtained pressure monitoring data of the long-distance pipeline, and m is a variable number;
2) Covariance matrix XX for calculating pressure monitoring data of long-distance pipeline T /(n-1)＝PΛP T Where Λ is a eigenvector of the long-distance pipeline pressure monitoring data matrix X, Λ=diag { λ } 1 ,λ 2 ,…,λ m Covariance matrix XX of pressure monitoring data of long-distance pipeline T /(n-1)＝PΛP T Feature vector Λ calculates variable P ε R m×m ；
3) Extracting a component matrix t=p from the long-distance pipeline pressure monitoring data matrix X according to the variable P T X；
4) Calculated variable q=Λ -1/2 P T Whitening the extracted component matrix T to obtain whitened result x ＝Λ -1/2 T＝Λ -1/2 P T X＝QX；
5) According to C T C=d conditional computation matrix C e R m×n Wherein d=diag { λ } 1 ,λ 2 ,…,λ d The feature vector of the d independent signals is known, and the matrix S=C is calculated according to the matrix C T Z；
6) According to C n T ＝D -1/2 C T C (C) n T C n Calculation matrix C =i n ；
7) Calculating a separation matrix W E R d×m Mixing matrix A epsilon R m×d Wherein, the method comprises the steps of, wherein,
8) Monitoring data matrix X epsilon R by separation matrix W m×n Signal separation is performed to obtain source signals of independent components, and the independent relation between the source signals of the independent components is reflected by non-Gaussian property, wherein the non-Gaussian property is formed by a negative entropy function J (W T X)＝[E{G(W T X)}-E{G( v )}] 2 Quantization is performed, wherein v is a Gaussian variable with zero mean and unit variance, and the negative entropy function can select three non-quadratic functions, wherein the three non-quadratic functions are respectivelyG 2 (u)＝exp(-a 2 u 2 2) and G 3 ＝u 4 Wherein 1.ltoreq.a 1 ≤2，a 2 About 1, the cosh () is a hyperbolic cosine function;
9) Three component importance evaluation criteria are constructed, wherein the three component importance evaluation criteria are respectively EMICA-CPV,EMICA-nG;
10 According to three non-quadratic functions and three component importance evaluation criteria, forming a double-layer comprehensive learning strategy, namelyWherein i epsilon {1,2,3}, i epsilon {1,2,3} represent three non-quadratic functions, E epsilon R, respectively m×n For the initially set residual matrix +.>The ith, jth mixing matrix, W, is the element of the ith row and jth column in the mixing matrix A i j To separate the first of the matrices Wi row j column elements;
11 Various non-quadratic functions to respectively select the above three component importance evaluation criteria to form 9 component selection modelsWhere i ε {1,2,3}, i ε {1,2,3}. Then, multiple statistics are formed into a unique index in a probability mode by Bayes reasoning;
12 In the ith component selection model, a monitoring statistic T is calculated 2 Is a detection probability of (2);
13 In the ith component selection model, calculating the detection probability of the statistic Q;
14 Separation to obtain abnormal component s i E S, and separate anomaly component S i Denoted as y i Recalculating the minimization loss functionObtaining a weight coefficient w, wherein x i As the true value of the outlier component, lambda is the determined regularization parameter and, I.I. | F Representing the Frobenius specification;
15 Using kernel transformation method to expand dominant regression function, and reusing Gaussian kernel function and polynomial functionObtaining regression fault signal data y, wherein sigma is a kernel parameter of a set Gaussian kernel function, and u and d are kernel parameters of a set polynomial kernel function respectively;
16 The transmission speed v of the pressure wave is calculated according to the returned fault signal data y, and the leakage position d is calculated according to the transmission speed v of the pressure wave.
Monitoring statistics T 2 Is of the detection probability of (2)The method comprises the following steps:
let N and F denote normal operating conditions and abnormal operating conditions respectively,a kind of electronic device with high-pressure air-conditioning systemConfidence degrees alpha and 1-alpha of calculated control limits are respectively expressed as the probability of normal operation condition N and abnormal operation condition FA kind of electronic device with high-pressure air-conditioning systemWherein, the liquid crystal display device comprises a liquid crystal display device,
the component selection models are synthesized as follows:
leak locationL pressure monitoring points are separated by two sections, delta t is the time difference of data receiving of the sensors at the head end and the tail end, and u is the fluid speed in the pipe.
The invention has the following beneficial effects:
according to the integrated improved ICA-KRR algorithm-based long-distance pipeline pressure monitoring method, when the integrated improved ICA-KRR algorithm-based long-distance pipeline pressure monitoring method is specifically operated, a long-distance pipeline pressure monitoring data matrix is constructed through collected pressure monitoring data, a component matrix T is extracted from the long-distance pipeline pressure monitoring data matrix, whitening is carried out on the component matrix, abnormal components are separated, three non-quadratic functions are combined with three component importance evaluation criteria, regression separation is carried out on pressure fault components based on Bayesian reasoning and dominant regression functions, fault signal data are determined, and then the leakage position is calculated according to the fault signal data, so that real-time monitoring and accurate positioning of the leakage position on the long-distance pipeline are achieved, workers are reminded of timely maintaining the pipeline, and unnecessary loss is avoided.
Drawings
FIG. 1 is a flow chart of the present invention;
FIG. 2 is a schematic diagram of source variables in a numerical simulation process;
FIG. 3 is a schematic diagram of individual component analysis monitoring details based on comprehensive improvement;
fig. 4 is a graph of average fault detection rate broken lines for different numbers of dominant ICs.
Detailed Description
The invention is described in further detail below with reference to the attached drawing figures:
the invention relates to a long-distance pipeline pressure monitoring method based on an integrated and improved ICA-KRR algorithm, which comprises the following steps:
1) Collecting pressure monitoring data of a plurality of long-distance pipelines, and constructing a long-distance pipeline pressure monitoring data matrix X epsilon R by the obtained pressure monitoring data of the long-distance pipelines m×n N is the number of obtained pressure monitoring data of the long-distance pipeline, and m is a variable number;
2) Covariance matrix XX for calculating pressure monitoring data of long-distance pipeline T /(n-1)＝PΛP T Where Λ is a eigenvector of the long-distance pipeline pressure monitoring data matrix X, Λ=diag { λ } 1 ,λ 2 ,…,λ m Covariance matrix XX of pressure monitoring data of long-distance pipeline T /(n-1)＝PΛP T Feature vector Λ calculates variable P ε R m×m ；
3) Extracting a component matrix t=p from the long-distance pipeline pressure monitoring data matrix X according to the variable P T X；
4) Calculated variable q=Λ -1/2 P T Whitening the extracted component matrix T to obtain whitened productPost result z=Λ -1/2 T＝Λ -1/2 P T X＝QX；
5) According to C T C=d conditional computation matrix C e R m×n Wherein d=diag { λ } 1 ,λ 2 ,…,λ d The feature vector of the d independent signals is known, and the matrix S=C is calculated according to the matrix C T Z ；
6) According to C n T ＝D -1/2 C T C (C) n T C n Calculation matrix C =i n ；
7) Calculating a separation matrix W E R d×m Mixing matrix A epsilon R m×d Wherein, the method comprises the steps of, wherein,
8) Monitoring data matrix X epsilon R by separation matrix W m×n Signal separation is performed to obtain source signals of independent components, and the independent relation between the source signals of the independent components is reflected by non-Gaussian property, wherein the non-Gaussian property is formed by a negative entropy function J (W T X)＝[E{G(W T X)}-E{G(v)}] 2 Quantization is performed, wherein v is a Gaussian variable with zero mean and unit variance, and the negative entropy function can select three non-quadratic functions, wherein the three non-quadratic functions are respectivelyG 2 (u)＝exp(-a 2 u 2 2) and G 3 ＝u 4 Wherein 1.ltoreq.a 1 ≤2，a 2 About 1, the cosh () is a hyperbolic cosine function;
9) Three component importance evaluation criteria are constructed, wherein the three component importance evaluation criteria are EMICA-CPV and EMICA-L respectively 2 EMICA-nG;
10 According to three non-quadratic functions and three component importance evaluation criteria, forming a double-layer comprehensive learning strategy, namelyWherein i is {1,2,3}, i is {1,2 }3 represents three non-quadratic functions, E.epsilon.R m×n For the initially set residual matrix +.>The ith, jth mixing matrix, W, is the element of the ith row and jth column in the mixing matrix A i j For separating elements of the ith row and jth column of the matrix W;
11 Various non-quadratic functions to respectively select the above three component importance evaluation criteria to form 9 component selection modelsWhere i ε {1,2,3}, i ε {1,2,3}. Then, multiple statistics are formed into a unique index in a probability mode by Bayes reasoning;
12 In the ith component selection model, a monitoring statistic T is calculated 2 Is a detection probability of (2);
13 In the ith component selection model, calculating the detection probability of the statistic Q;
14 Separation to obtain abnormal component s i E S, and separate anomaly component S i Is recorded as yi Recalculating the minimization loss functionObtaining a weight coefficient w, wherein x i As the true value of the outlier component, lambda is the determined regularization parameter and, I.I. | F Representing the Frobenius specification;
15 Using kernel transformation method to expand dominant regression function, and reusing Gaussian kernel function and polynomial functionK (y, y) i )＝(uyy i +1) d Obtaining regression fault signal data y, wherein sigma is a kernel parameter of a set Gaussian kernel function, and u and d are kernel parameters of a set polynomial kernel function respectively;
16 The transmission speed v of the pressure wave is calculated according to the returned fault signal data y, and the leakage position d is calculated according to the transmission speed v of the pressure wave.
Monitoring statistics T 2 Is of the detection probability of (2)The method comprises the following steps:
let N and F denote normal operating conditions and abnormal operating conditions respectively,a kind of electronic device with high-pressure air-conditioning systemConfidence a and 1 of the calculated control limits respectively - The probabilities of the normal operation condition N and the abnormal operation condition F are respectively expressed asA kind of electronic device with high-pressure air-conditioning systemWherein, the liquid crystal display device comprises a liquid crystal display device,
the component selection models are synthesized as follows:
leak locationL pressure monitoring points are separated by two sections, delta t is the time difference of data receiving of the sensors at the head end and the tail end, and u is the fluid speed in the pipe.
Simulation experiment
The simulation experiment carries out numerical simulation experiment analysis on a TE platform, and firstly, the MICA model is utilized to realize TE process fault detection; and then, the constructed KRR regression model is utilized to further carry out the selection and separation of fault variables on fault data, and finally, the fault positioning and diagnosis are realized.
The physical model of TE process consists of reactor, condenser, gas-liquid separator, circulating compressor and stripping tower. The TE process consists of 41 measured variables (22 continuous variables and 19 component value variables) and 12 manipulated variables. The simulation experiment selects 30 variables, including 10 operating variables and 20 measured variables, wherein 20 measured variables are shown in table 1. There are 21 programmable exception conditions in the TE process; compared with a method using different non-quadratic functions, the recognition degree of the dominant IC is higher.
TABLE 1
In the actual long-distance pipeline running process, most faults can cause two or more variable anomalies, so that more targeted inspection can be performed according to main anomalies; in the real-time monitoring process, the average fault detection rates of different numbers of dominant components are compared, so that the invention has simple and visual variable selection capability.
What is not described in detail in the present specification belongs to the prior art that is well known to those skilled in the art.
The above embodiments are merely illustrative of the present invention and are not intended to be limiting. Although the related embodiments of the present invention and the accompanying drawings are disclosed for illustrative purposes, those skilled in the art will appreciate; various alternatives, variations, and modifications are possible without departing from the spirit and scope of the invention and the appended claims. Therefore, all equivalent solutions are intended to be within the scope of the present invention, which is to be defined by the claims and not limited to the preferred embodiments and the disclosure of the accompanying drawings.
Claims (1)
1. The long-distance pipeline pressure monitoring method based on the integrated and improved ICA-KRR algorithm is characterized by comprising the following steps of:
1) Collecting pressure monitoring data of a plurality of long-distance pipelines, and constructing a long-distance pipeline pressure monitoring data matrix X epsilon R by the obtained pressure monitoring data of the long-distance pipelines m×n N is the number of obtained pressure monitoring data of the long-distance pipeline, and m is a variable number;
2) Covariance matrix XX for calculating pressure monitoring data of long-distance pipeline T /(n-1)＝PΛP T Where Λ is a eigenvector of the long-distance pipeline pressure monitoring data matrix X, Λ=diag { λ } 1 ,λ 2 ,…,λ m Covariance matrix XX of pressure monitoring data of long-distance pipeline T /(n-1)＝PΛP T Feature vector Λ calculates variable P ε R m×m ；
3) Extracting a component matrix t=p from the long-distance pipeline pressure monitoring data matrix X according to the variable P T X；
4) Calculating statistics q=Λ -1/2 P T Whitening the extracted component matrix T to obtain whitened result Z=Λ -1/2 T＝Λ -1/2 P T X＝QX；
5) According to C T C=d conditional computation matrix C e R m×n Wherein d=diag { λ } 1 ,λ 2 ,…,λ d The feature vector of the d independent signals is known, and the matrix S=C is calculated according to the matrix C T Z；
6) According to C n T ＝D -1/2 C T C (C) n T C n Calculation matrix C =i n ；
7) Calculating a separation matrix W E R d×m Mixing matrix A epsilon R m×d Wherein, the method comprises the steps of, wherein,
8) Monitoring data matrix X epsilon R by separation matrix W m×n Signal separation is performed to obtain source signals of independent components, and the independent relation between the source signals of the independent components is reflected by non-Gaussian property, wherein the non-Gaussian property is formed by a negative entropy function J (W T X)＝[E{G(W T X)}-E{G(v)}] 2 Quantization is performed, wherein v is a Gaussian variable with zero mean and unit variance, and the negative entropy function can select three non-quadratic functions, wherein the three non-quadratic functions are respectivelyG 2 (u)＝exp(-a 2 u 2 2) and G 3 ＝u 4 Wherein 1.ltoreq.a 1 ≤2，a 2 About 1, the cosh () is a hyperbolic cosine function;
9) Three component importance evaluation criteria are constructed, wherein the three component importance evaluation criteria are EMICA-CPV and EMICA-L respectively 2 EMICA-nG;
10 According to three non-quadratic functions and three component importance evaluation criteria, forming a double-layer comprehensive learning strategy, namelyWherein i epsilon {1,2,3}, i epsilon {1,2,3} represent three non-quadratic functions, H epsilon R, respectively m×n For the initially set residual matrix, A i j The ith, jth mixing matrix, W, is the element of the ith row and jth column in the mixing matrix A i j For separating elements of the ith row and jth column of the matrix W;
11 Various non-quadratic functions to respectively select the above three component importance evaluation criteria to form 9 component selection modelsWherein, i epsilon {1,2,3}, and then forming a unique index by using Bayesian reasoning to form multiple statistics in a probability manner;
12 At the ith component selection modelIn calculating the monitoring statistic T 2 Is a detection probability of (2);
13 In the ith component selection model, calculating the detection probability of the statistic Q;
14 Separation to obtain abnormal component s i E S, and separate anomaly component S i Denoted as y i Recalculating the minimization loss functionObtaining a weight coefficient w, wherein x i As the true value of the outlier component, lambda is the determined regularization parameter and, I.I. | F Representing the Frobenius specification;
15 Using kernel transformation method to expand dominant regression function, and reusing Gaussian kernel function and polynomial functionK (y, y) i )＝(cyy i +1) e Obtaining regression fault signal data y, wherein sigma is a kernel parameter of a set Gaussian kernel function, and c and e are kernel parameters of a set polynomial kernel function respectively;
16 Calculating the transmission speed of the pressure wave according to the returned fault signal data y, and calculating the leakage position d according to the transmission speed of the pressure wave;
monitoring statistics T 2 Is of the detection probability of (2)The method comprises the following steps:
let N and F denote normal operating conditions and abnormal operating conditions respectively,is->Confidence degrees alpha and 1-alpha of calculated control limit are respectively expressed as +.o.>Is->Wherein (1)>
The component selection models are synthesized as follows:
leak locationL pressure monitoring points are separated by two sections, delta t is the time difference of data receiving of the sensors at the head end and the tail end, z is the fluid speed in the pipe, and L is the transmission speed of pressure waves.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810285077.6A CN108804740B (en) | 2018-04-02 | 2018-04-02 | Long-distance pipeline pressure monitoring method based on integrated improved ICA-KRR algorithm |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810285077.6A CN108804740B (en) | 2018-04-02 | 2018-04-02 | Long-distance pipeline pressure monitoring method based on integrated improved ICA-KRR algorithm |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN108804740A CN108804740A (en) | 2018-11-13 |
| CN108804740B true CN108804740B (en) | 2023-09-29 |
Family
ID=64095523
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810285077.6A Active CN108804740B (en) | 2018-04-02 | 2018-04-02 | Long-distance pipeline pressure monitoring method based on integrated improved ICA-KRR algorithm |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108804740B (en) |
Families Citing this family (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109442221B (en) * | 2018-11-21 | 2019-10-18 | 同济大学 | A kind of water supply network booster method for detecting extracted based on pressure disturbance |
| CN111898313B (en) * | 2020-06-30 | 2022-05-20 | 电子科技大学 | Fault detection method based on ICA and SVM integrated learning |
| CN112488371B (en) * | 2020-11-20 | 2022-09-13 | 国网福建省电力有限公司厦门供电公司 | Pipeline intelligent early warning method and system based on big data |
| CN113551158B (en) * | 2021-07-14 | 2022-06-07 | 吉林大学 | Method for determining quantity and position of pipeline leakage based on Bayesian framework |
| CN113864665B (en) * | 2021-10-09 | 2024-02-13 | 重庆邮电大学 | Fluid pipeline leakage positioning method based on adaptive ICA and improved RLS filter |
| CN115130496B (en) * | 2022-05-24 | 2023-06-16 | 北京化工大学 | Pipeline pressure signal abnormality detection method based on Bagging and RM-LOF integrated single classifier |
| CN115307670B (en) * | 2022-10-12 | 2023-01-13 | 蘑菇物联技术（深圳）有限公司 | Method, apparatus and medium for locating central air conditioning system anomaly sensors |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104180166A (en) * | 2014-07-09 | 2014-12-03 | 中国石油大学(华东) | Pipeline leakage detection method based on pipeline pressure data |
| CN106290578A (en) * | 2016-07-27 | 2017-01-04 | 常州大学 | The detection of a kind of pressure pipeline Small leak source and accurate positioning method |
| CN106372299A (en) * | 2016-08-29 | 2017-02-01 | 西安建筑科技大学 | Pipeline remaining life predication method based on inverse Gaussian process and EM-PF |
| CN106599564A (en) * | 2016-12-07 | 2017-04-26 | 河海大学常州校区 | Failure monitoring and self-healing method for multi-agent combustion gas pipeline based on immune learning |
-
2018
- 2018-04-02 CN CN201810285077.6A patent/CN108804740B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104180166A (en) * | 2014-07-09 | 2014-12-03 | 中国石油大学(华东) | Pipeline leakage detection method based on pipeline pressure data |
| CN106290578A (en) * | 2016-07-27 | 2017-01-04 | 常州大学 | The detection of a kind of pressure pipeline Small leak source and accurate positioning method |
| CN106372299A (en) * | 2016-08-29 | 2017-02-01 | 西安建筑科技大学 | Pipeline remaining life predication method based on inverse Gaussian process and EM-PF |
| CN106599564A (en) * | 2016-12-07 | 2017-04-26 | 河海大学常州校区 | Failure monitoring and self-healing method for multi-agent combustion gas pipeline based on immune learning |
Non-Patent Citations (7)
| Title |
|---|
| Chudong Tong 等.Double-layer ensemble monitoring of non-gaussian processes using modified independent component analysis.《ISA Transactions》.2017,第68卷第181-188页. * |
| Chudong Tong 等.Ensemble modified independent component analysis for enhanced non-Gaussian process monitoring.《Control Engineering Practice》.2016,第58卷第34-41页. * |
| wavesCuiwei Liu 等.Leakage monitoring research and design for natural gas pipelinesbased on dynamic pressure waves.《Journal of Process Control》.2016,第50卷第66-76页. * |
| 刘剑 等.基于Multi-kernel和KRR的数据还原算法.《控制与决策》.2014,第29卷(第5期),第821-826页. * |
| 王海蓉 等.基于支持向量机的天然气管道小泄漏信号检测.《石油机械》.2008,第36卷(第12期),第43-45,86页. * |
| 衷路生 等.基于ICA-PCA和Lasso的过程故障诊断.《华中科技大学学报(自然科学版)》.2016,第44卷(第10期),第98-102页. * |
| 辜小花等.基于动态核独立元统计量的石油管道泄漏检测.《仪器仪表学报》.2017,第38卷(第01期),第166-173页. * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN108804740A (en) | 2018-11-13 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108804740B (en) | Long-distance pipeline pressure monitoring method based on integrated improved ICA-KRR algorithm | |
| CN108426713B (en) | Rolling bearing weak fault diagnosis method based on wavelet transformation and deep learning | |
| CN108931387B (en) | Fault diagnosis method based on multi-sensor signal analysis and capable of providing accurate diagnosis decision | |
| CN105240187B (en) | The method that turbine state monitoring and fault diagnosis are realized based on chaology | |
| CN111259864B (en) | Method for identifying running state of water turbine | |
| CN105834835A (en) | Method for monitoring tool wear on line based on multiscale principal component analysis | |
| WO2019080367A1 (en) | Method for evaluating health status of mechanical device | |
| CN112016251B (en) | Nuclear power device fault diagnosis method and system | |
| CN111853555A (en) | Water supply pipe network blind leakage identification method based on dynamic process | |
| CN105607631B (en) | The weak fault model control limit method for building up of batch process and weak fault monitoring method | |
| Harmouche et al. | Long-term monitoring for leaks in water distribution networks using association rules mining | |
| CN103742794A (en) | Simulating device and simulating method for pipeline leakage acoustic emission signals | |
| CN117407827B (en) | Abnormal operation data detection method for purification engineering waste gas purification equipment | |
| CN113822201B (en) | Deep learning method for underwater object shape recognition based on flow field velocity component time course | |
| CN115824519B (en) | Comprehensive diagnosis method for valve leakage faults based on multi-sensor information fusion | |
| CN203743867U (en) | Pipeline leakage acoustic emission signal simulation device | |
| CN111695465A (en) | Pipe network fault diagnosis and positioning method and system based on pressure wave mode identification | |
| Zhang et al. | Fault diagnosis of natural gas pipeline leakage based on 1D-CNN and self-attention mechanism | |
| CN114444663A (en) | Water supply pipe network leakage detection and positioning method based on time convolution network | |
| Lang et al. | Pipeline leak aperture recognition based on wavelet packet analysis and a deep belief network with ICR | |
| CN103175687A (en) | Fault location method for sliding-tooth reducer | |
| CN113298133B (en) | Method for diagnosing explosion tube fault of supercritical unit boiler | |
| CN115293188A (en) | Fault diagnosis method and device for reciprocating mechanical equipment | |
| CN109946096B (en) | High-speed train air pipe fault diagnosis method based on model space | |
| CN110516323A (en) | A kind of pressure pipeline damage forecast method based on Time-Series analysis |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |