A method and device for identifying the cause of abnormal operation of a transmission pipeline

Abstract

The embodiments of this specification provide a method and device for identifying the cause of abnormal operation of a transmission pipeline. The method includes: obtaining operation data of different positions of the transmission pipeline; clustering the operation data of the different positions to obtain a preset number of operation data with different abnormal fluctuation types; establishing labels for the preset number of operation data with different abnormal fluctuation types according to the correspondence between the abnormal fluctuation type and the abnormal cause; the label is the abnormal cause corresponding to the abnormal fluctuation type of the operation data; using the operation data with the label as a training sample, calculating the classification conditions of the operation data in the training sample; identifying the abnormal cause of the operation data to be identified according to the classification condition, thereby improving the accuracy of identifying the cause of abnormal operation of the transmission pipeline.

Description

Technical Field

[0001] The embodiments of this specification relate to the field of oil transportation technology, and more particularly to a method and device for identifying the cause of abnormal operation of a transportation pipeline. Background Art

[0002] With the development of the oil and gas industry, pipeline transportation has become another important means of transportation after railways, roads, waterways and air transportation. Pipeline transportation plays an increasingly important role in the transportation of oil, natural gas and other fluids. Pipelines are used as carriers for the transmission of liquids such as oil, and are laid in various environments such as land and sea around the world. Pipeline transportation is more convenient, but it is easy to cause damage to oil pipelines after long-term use during pipeline transportation, which will cause oil loss and waste resources. Once a leak occurs in an oil pipeline, it will not only bring significant economic losses and environmental pollution problems, but also affect production progress, and may also affect the safety of the oil pipeline and cause personal injury.

[0003] At present, pipeline fault diagnosis is mainly divided into direct method and indirect method. The direct method is to use human senses, such as direct observation, knocking, etc., or some special devices, such as pipeline robots, to detect the pipeline and the environment in which the pipeline is located, so as to determine whether the pipeline has a fault. The indirect method mainly uses the changes in the pressure, flow, temperature and other data in the pipeline provided by the data acquisition system, and uses a specific algorithm to analyze and process the detection data to realize the fault judgment.

[0004] However, data collected by indirect methods generally have data uncertainties, such as measurement noise, transmission noise, sampling errors, and observation blind spots. It is impossible to intuitively detect the data at various points along the pipeline, which often leads to large deviations between the measured data and the standard theoretical analysis results.

[0005] In the prior art, the direct method requires human experience to judge the abnormality of the pipeline, which is prone to misjudgment or inaccuracy. The data collected by the indirect method is uncertain, and the processing of large amounts of data is time-consuming, and the real-time online detection capability of faults is limited. Summary of the invention

[0006] The purpose of the embodiments of this specification is to provide a method and device for identifying the cause of abnormal operation of a transportation pipeline, so as to improve the accuracy of identifying the cause of abnormal operation of a transportation pipeline.

[0007] To solve the above problems, an embodiment of the present specification provides a method for identifying the cause of abnormal operation of a transmission pipeline, the method comprising: acquiring operation data of different positions of the transmission pipeline; clustering the operation data of the different positions to obtain a preset number of operation data with different abnormal fluctuation types; establishing labels for the preset number of operation data with different abnormal fluctuation types according to the correspondence between the abnormal fluctuation type and the abnormal cause; the label is the abnormal cause corresponding to the abnormal fluctuation type appearing in the operation data; using the operation data with the label as a training sample, calculating the classification conditions of the operation data in the training sample; and identifying the abnormal cause of the operation data to be identified according to the classification condition.

[0008] To solve the above problems, an embodiment of the present specification also provides a device for identifying the cause of abnormal operation of a transmission pipeline, the device comprising: an acquisition module, used to acquire operation data of different positions of the transmission pipeline; a clustering module, used to cluster the operation data of the different positions to obtain a preset number of operation data with different abnormal fluctuation types; an establishment module, used to establish labels for the preset number of operation data with different abnormal fluctuation types according to the correspondence between the abnormal fluctuation type and the abnormal cause; the label is the abnormal cause corresponding to the abnormal fluctuation type appearing in the operation data; a calculation module, used to use the operation data with the label as a training sample, and calculate the classification conditions of the operation data in the training sample; an identification module, used to identify the abnormal cause of the operation data to be identified according to the classification condition.

[0009] It can be seen from the technical solutions provided in the above embodiments of this specification that in the embodiments of this specification, the operation data of different positions of the transmission pipeline can be obtained; the operation data of the different positions are clustered to obtain a preset number of operation data with different abnormal fluctuation types; labels are established for the preset number of operation data with different abnormal fluctuation types according to the correspondence between the abnormal fluctuation type and the abnormal cause; the label is the abnormal cause corresponding to the abnormal fluctuation type of the operation data; the operation data with the label is used as a training sample, and the classification conditions of the operation data in the training sample are calculated; the abnormal cause of the operation data to be identified is identified according to the classification condition. The method provided in the embodiments of this specification can realize the intelligent perception of abnormal fluctuations in the operation data during the operation of the pipeline, improve the accuracy of identifying the abnormal cause of the pipeline operation, and provide support for the operation decision of the pipeline. BRIEF DESCRIPTION OF THE DRAWINGS

[0010] In order to more clearly illustrate the embodiments of this specification or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in this specification. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0011] Figure 1 This is a flow chart of a method for identifying the cause of abnormal operation of a transmission pipeline according to an embodiment of this specification;

[0012] Figure 2 This is a diagram showing the results of identifying abnormal causes of a liquid delivery pipeline according to an embodiment of this specification;

[0013] Figure 3 This is a schematic diagram of the accuracy of identifying abnormal causes of liquid delivery pipelines according to the embodiments of this specification;

[0014] Figure 4 This is a schematic diagram of the functional modules of a device for identifying the cause of abnormal operation of a transportation pipeline according to an embodiment of this specification. DETAILED DESCRIPTION

[0015] The following will be combined with the drawings in the embodiments of this specification to clearly and completely describe the technical solutions in the embodiments of this specification. Obviously, the described embodiments are only part of the embodiments of this specification, not all of the embodiments. Based on the embodiments in this specification, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of this specification.

[0016] At present, pipeline fault diagnosis is mainly divided into direct method and indirect method. The direct method is to use human senses, such as direct observation, knocking, etc., or some special devices, such as pipeline robots, to detect the pipeline and the environment in which the pipeline is located, so as to determine whether the pipeline has a fault. The indirect method mainly uses the changes in the pressure, flow, temperature and other data in the pipeline provided by the data acquisition system, and uses a specific algorithm to analyze and process the detection data to realize the fault judgment. The existing operation monitoring technology of oil and other liquid pipelines has three main deficiencies:

[0017] (1) Manual tracking of the interface migration status is adopted, which has poor accuracy. Quality accidents have occurred on site due to untimely tracking, and the untimely and accurate download of oil products has affected the market supply. In addition, the inability to accurately track the oil product interface increases the difficulty of pipeline oil inventory and the difficulty of digital quality management.

[0018] (2) The current pipeline monitoring system has a blind spot in pressure observation and cannot observe the pressure at each point along the pipeline in real time and intuitively. It cannot effectively avoid the occurrence of high-point emptying and low-point overpressure, which causes pipeline corrosion, a large amount of oil mixing, pipe burst and other problems.

[0019] (3) Judging based on human experience: Processing equipment may malfunction during operation due to various reasons. Lack of scientificity and rigor, insufficient comprehensive analysis, and incorrect or inaccurate judgments may occur.

[0020] Considering that if data mining and artificial intelligence technology are introduced into the pipeline operation monitoring technology, combined with the actual operation monitoring of liquid pipelines such as oil, from batch tracking, flow pressure prediction, pre-perception of operation anomalies, online identification of operation anomalies and other aspects, the cause of abnormal operation of the pipeline can be identified, it is expected to solve the problem of easy misjudgment or inaccuracy in the existing technology for pipeline anomalies, improve the accuracy of pipeline operation anomaly identification, and provide support for pipeline operation decisions. Based on this, the embodiment of this specification provides a method and device for identifying the cause of abnormal operation of a pipeline.

[0021] Figure 1 This is a flow chart of a method for identifying the cause of abnormal operation of a transportation pipeline according to an embodiment of this specification. In the embodiment of this specification, the subject that executes the method for identifying the cause of abnormal operation of a transportation pipeline may be an electronic device with a logical operation function, and the electronic device may be a server or a client, and the client may be a desktop computer, a tablet computer, a laptop computer, a workstation, etc. Of course, the client is not limited to the above-mentioned electronic device with a certain entity, and it can also be software running in the above-mentioned electronic device. It can also be a program software formed by program development, and the program software can run in the above-mentioned electronic device. Figure 1 As shown, the method for identifying the cause of abnormal operation of a conveying pipeline may include the following steps.

[0022] S110: Acquire operation data of different positions of the transmission pipeline.

[0023] In some embodiments, the operation data may include pressure data or flow data. The operation data may be obtained by a pressure sensor or a flow sensor. Specifically, a plurality of pressure sensors or a plurality of flow sensors may be arranged at different positions in the delivery pipeline, and the pressure data or flow data at different positions of the delivery pipeline may be obtained according to the detection data of the pressure sensor or the flow sensor.

[0024] In some embodiments, the server may use any method to obtain the operation data of different positions of the delivery pipeline. For example, after reading the detection data of the pressure sensor or the flow sensor, the user may directly send the operation data of different positions of the delivery pipeline to the server, and the server may receive it; for example, other electronic devices other than the server may send the operation data of different positions of the delivery pipeline to the server, and the server may receive it. In the embodiments of this specification, the method used by the server to obtain the operation data of different positions of the delivery pipeline is not limited.

[0025] Some missing data may appear in the operation data obtained at different positions of the conveying pipeline. There are various reasons for the generation of missing data, which can be mainly divided into mechanical reasons and human reasons. Mechanical reasons are data missing caused by failure in data collection or preservation due to mechanical reasons, such as failure in data storage, memory damage, and mechanical failure resulting in failure to collect data for a certain period of time (for timed data collection). Human reasons are data missing due to subjective errors, historical limitations or intentional concealment. Based on this, in some embodiments, after obtaining the operation data at different positions of the conveying pipeline, the server can pre-process the obtained operation data at different positions of the conveying pipeline, and the pre-processing can include filling the missing data in the operation data by interpolation.

[0026] Specifically, the two-point cubic Hermite interpolation method can be used to construct a method for filling missing data in the continuously fluctuating operation data of liquid transportation pipelines such as oil. The process of filling missing values ​​of pipeline operation data by the two-point cubic Hermite interpolation method includes: let (x1, y1) be the missing data between the operation data (x0, y0) and (x2, y2), x0, x1, x2 are the position coordinates; y0, y1, y2 are the values ​​of the operation data. The two-point cubic Hermite interpolation method is used to fill in the pipeline operation data, and the formula is as follows:

[0027] y1=y0α0(x1)+y2α1(x1)+y'0β0(x1)+y'2β1(x1) (1)

[0028] Among them, α and β are interpolation basis functions, which are obtained by the missing point coordinate x1 and the adjacent point coordinates x0 and x2.

[0029] S120: Clustering the operation data at different locations to obtain a preset number of operation data with different abnormal fluctuation types.

[0030] In some embodiments, if the server performs a preprocessing step to fill in missing data in the acquired operation data at different locations of the transmission pipeline, then in S120, the preprocessed operation data can be clustered to obtain a preset number of operation data with different abnormal fluctuation types.

[0031] In some embodiments, the operation data at different locations can be clustered using phase space reconstruction technology, and a method for characterizing the fluctuation of the operation data of the pipeline can be established through the dynamic perspective of time series data to obtain a method for observing the fluctuation of the operation data of the pipeline; on this basis, a K-means model can be used to construct an abnormal fluctuation type discrimination model to form a method for discriminating abnormal fluctuations of the operation data of the pipeline, thereby avoiding detection errors caused by a single identification method. Based on this, clustering the operation data at different locations to obtain a preset number of operation data with different abnormal fluctuation types can include the following steps.

[0032] S121: Reconstruct the operating data in phase space, map the operating data into vectors in phase space, and obtain phase space reconstructed data.

[0033] The state of a system at a certain moment is called a phase, and the geometric space that determines the state is called a phase space. Generally speaking, the phase space of a nonlinear system may have a very high dimension, even infinite, and sometimes the dimension is unknown. In order to obtain qualitative information of a dynamic system, it is often necessary to know sufficient state evolution information, but in actual engineering applications, data acquisition equipment can generally only obtain a one-dimensional vector that reflects system information, that is, a time series. According to Takens embedding theorem: For infinitely long, noise-free one-dimensional chaotic attractor, an n-dimensional embedded phase space can be found in the sense of topological invariance. In the embodiments of this specification, the operating data obtained is a time series. According to Takens embedding theorem, the qualitative information of a dynamic system can be extracted from a time series through phase space reconstruction technology.

[0034] Specifically, the implementation process of the phase space reconstruction technology includes the following input data: pipeline operation data in the form of time series: {y1, y2, ..., y n}; Based on Takens embedding theorem, the time series data of pipeline operation parameters {y1, y2, …, y n} is mapped to a vector in the d-dimensional phase space to simplify the representation of complex operating parameter data. The mapped phase space vector is as follows:

[0035] {y1, y2, ..., y n}={z n ,...,z n+(d-1)τ} (2)

[0036] Among them, y1, y2, ..., y n is the value of the running data at different locations; n ,...,z n+(d-1)τ is the phase space vector after mapping, i.e., the phase space reconstruction data; d represents the dimension of the phase space; τ represents the delay time.

[0037] Phase space reconstruction technology has two key parameters: embedding dimension and delay time. In Takens embedding theorem, embedding dimension and delay time are only theoretically proved to exist, and no specific expression is given. In practical applications, time series are finite series with noise. Embedding dimension and delay time must be selected according to the actual situation. Methods for obtaining delay time can include autocorrelation method, mean shift method, complex autocorrelation method and mutual information method. Methods for obtaining embedding dimension can include geometric invariant method, false nearest neighbor method, etc.

[0038] S122: Clustering the phase space reconstruction data using a K-means clustering algorithm to obtain a preset number of operating data with different abnormal fluctuation types.

[0039] In some embodiments, the K-means clustering algorithm is an iterative clustering analysis algorithm, the steps of which are to pre-divide the data into K groups, then randomly select K objects as initial cluster centers, and then calculate the distance between each object and each seed cluster center, and assign each object to the cluster center closest to it. The cluster centers and the objects assigned to them represent a cluster. Each time a sample is assigned, the cluster center of the cluster is recalculated based on the existing objects in the cluster. This process will be repeated until a certain termination condition is met. The termination condition can be that no (or a minimum number) of objects are reassigned to different clusters, no (or a minimum number) of cluster centers change again, and the sum of squared errors is locally minimized.

[0040] In the embodiment of the present specification, the process of clustering the phase space reconstruction data using the K-means clustering algorithm to obtain a preset number of operating data with different abnormal fluctuation types includes the following steps.

[0041] The input data includes: the sample set is D = {z n ,...,z n+(d-1)τ}, the number of clusters is k, and the maximum number of iterations is N.

[0042] The output data includes: clusters are divided into C = {C1, C2, ..., C k}. Among them, C represents the abnormal fluctuation type, and different subscripts represent different abnormal fluctuation types.

[0043] Specifically, k samples can be randomly selected from the data set D as the initial k centroid vectors: {μ1,μ2,...,μ3}; for i = 1,2,...,N, the cluster C is initialized as For n = 1, 2, ..., m, calculate sample Z n and each centroid vector μ j The distance of (j=1,2,...,k) is calculated as follows:

[0044]

[0045] Z n The distance d obtained by assigning to the nearest centroid nj The corresponding category is λ n , the updated cluster is:

[0046]

[0047] For j = 1, 2, ..., k, for C j Recalculate the new centroid of all sample points in the calculation method:

[0048]

[0049] If all k centroid vectors have not changed, the output cluster is divided into C = {C1, C2, ..., C k}, different clusters represent different abnormal fluctuation types.

[0050] S130: Creating labels for the preset number of operation data with different abnormal fluctuation types according to the correspondence between the abnormal fluctuation type and the abnormal cause; the label is the abnormal cause corresponding to the abnormal fluctuation type appearing in the operation data.

[0051] In some embodiments, different abnormal fluctuation types may correspond to different abnormal causes. The abnormal causes corresponding to different abnormal fluctuation types may be predetermined by the staff. The abnormal causes may include abnormalities caused by pump shutdown, abnormalities caused by voltage fluctuations of the oil delivery pump, abnormalities caused by abnormal valve closing, etc.

[0052] In some embodiments, the server may create labels for operating data of different abnormal fluctuation types, where the labels are abnormal causes corresponding to the abnormal fluctuation types appearing in the operating data.

[0053] S140: Using the operation data with labels as training samples, calculating classification conditions of the operation data in the training samples.

[0054] In some embodiments, the operation data with labels can be used as training samples, and the classification conditions of the operation data in the training samples can be calculated by a machine learning algorithm. The calculation of the classification conditions of the operation data in the training samples can include the following steps.

[0055] S141: Extracting the eigenvalue vector of the running data in the training sample.

[0056] The eigenvalue vector of the operating data may be phase space reconstructed data obtained by performing phase space reconstruction on the operating data in S121 and mapping the operating data into a vector in the phase space.

[0057] S142: Based on the extracted eigenvalue vector of each running data in the training sample, a machine learning algorithm is used to calculate the classification conditions of the running data in the training sample.

[0058] For operation data of different abnormal fluctuation types, the classification conditions of the operation data in the training sample can be calculated based on the extracted eigenvalue vector of each operation data in the training sample. For example, the eigenvalue vectors of the operation data of the same abnormal fluctuation type are similar, and the eigenvalue vectors of the operation data of different abnormal fluctuation types are quite different. Therefore, the classification conditions of the operation data in the training sample can be calculated based on the extracted eigenvalue vector of each operation data in the training sample using a machine learning algorithm.

[0059] In some embodiments, a machine algorithm may be used to calculate the classification conditions of the running data in the training sample. Specifically, a machine learning algorithm may be used to construct a classification model. For example, a logistic regression algorithm, a decision tree algorithm, or a support vector machine algorithm may be used to construct a classification model. Of course, any other machine learning algorithm may also be used to construct a classification model, and this specification embodiment does not limit this.

[0060] In some embodiments, the eigenvalue vector of each running data in the training sample extracted can be used as input to train the classification model, and the classification model can calculate the classification conditions of the running data in the training sample based on the eigenvalue vector of each running data in the training sample extracted.

[0061] Here, the implementation method of calculating the classification conditions of the running data in the training sample is introduced by taking the support vector machine (SVM) algorithm as an example.

[0062] A nonlinear SVM model is established. Specifically, for the kernel function K(x,z), there is a mapping φ(x) from the input space to the feature space. For any point x,z in the input space, the kernel function can be expressed as:

[0063] K(x,z)=φ(x)·φ(z) (6)

[0064] Among them, φ(x)·φ(z) is the inner product of φ(x) and φ(z), and φ(x) and φ(z) are mapping functions.

[0065] In the dual problem of linear SVM, the kernel function K(x,z) is used to replace the inner product and the nonlinear SVM model is established as follows:

[0066]

[0067] Where i=1,2,...,N represents the slack variable of each training sample point; Represents the weight b * Indicates the threshold value.

[0068] The principle of nonlinear SVM parameter training is as follows:

[0069] Input: training data set T = {(x1,y1),(x2,y2),...,(x N ,y N )}, where x i ∈R n ,y i ∈{+1,-1}, i=1,2,...,N; Output: separating hyperplane and classification decision function.

[0070] Select appropriate kernel function K(x,z) and penalty parameter C>0, construct and solve the convex quadratic programming problem. The objective function is:

[0071]

[0072] The constraints are:

[0073]

[0074] The optimal solution is:

[0075]

[0076] Select α * A component of Satisfy the conditions Calculate b * The calculation formula is:

[0077]

[0078] Finally, the classification decision function is calculated. Because the Gaussian kernel function The corresponding SVM is a Gaussian radial basis function classifier, so in this case, the final classification decision function is obtained, that is, the classification condition is:

[0079]

[0080] Among them, σ represents the width parameter, which is used to control the radial range of the kernel function.

[0081] S150: Identify the abnormal cause of the operating data to be identified according to the classification condition.

[0082] In some embodiments, identifying the abnormal cause of the operating data to be identified according to the classification condition may include: extracting the eigenvalue vector of the operating data to be identified; calculating the classification value of the eigenvalue vector of the operating data to be identified according to the classification condition; and determining the abnormal cause of the operating data to be identified according to the classification value.

[0083] Specifically, for extracting the eigenvalue vector of the operating data to be identified, reference can be made to S121, where the eigenvalue vector of the operating data to be identified is the phase space reconstruction data obtained after phase space reconstruction. When the eigenvalue vector of the operating data to be identified is substituted into the classification condition, the classification value of the eigenvalue vector of the operating data to be identified can be obtained. Different classification values ​​can correspond to different abnormal causes. For example, if the classification value of the eigenvalue vector of the operating data to be identified is 1, the abnormal cause may be an abnormality caused by pump shutdown; if the classification value of the eigenvalue vector of the operating data to be identified is 2, the abnormal cause may be an abnormality caused by voltage fluctuation of the oil pump; or if the classification value of the eigenvalue vector of the operating data to be identified is in the interval [1,5], the abnormal cause may be an abnormality caused by pump shutdown; if the classification value of the eigenvalue vector of the operating data to be identified is in the interval [5,10], the abnormal cause may be an abnormality caused by voltage fluctuation of the oil pump. Of course, the above is only an illustrative example of determining the abnormal cause of the operating data to be identified based on the classification value. In the embodiments of this specification, the corresponding relationship between the classification value and the abnormal cause can also be set according to the actual situation, so as to determine the abnormal cause of the operating data to be identified based on the classification value.

[0084] The method for identifying the cause of abnormal operation of a transmission pipeline provided in the embodiments of this specification can obtain the operation data of different positions of the transmission pipeline; cluster the operation data of the different positions to obtain a preset number of operation data with different abnormal fluctuation types; establish labels for the preset number of operation data with different abnormal fluctuation types according to the correspondence between the abnormal fluctuation type and the abnormal cause; the label is the abnormal cause corresponding to the abnormal fluctuation type of the operation data; use the operation data with the label as a training sample, calculate the classification conditions of the operation data in the training sample; and identify the abnormal cause of the operation data to be identified according to the classification conditions. The method provided in the embodiments of this specification can realize the intelligent perception of abnormal fluctuations in the operation data during the operation of the pipeline, improve the accuracy of identifying the cause of abnormal operation of the pipeline, and provide support for the operation decision of the pipeline.

[0085] Combine the following Figure 2 and Figure 3 A specific embodiment of the method for identifying the cause of abnormal operation of a transmission pipeline provided in the embodiments of this specification is described.

[0086] A transportation model of a liquid pipeline is established based on the TLnet software under Pipeline Studio. The diameter of the pipeline is Φ400×6mm, the length of the pipeline is 60km, and there is one pump station and two shut-off valves. The pipeline operation is simulated by the TLnet simulation software to generate operating data such as pressure and flow. At the same time, in addition to normal operating conditions, multiple abnormal causes such as oil pump voltage fluctuations, abnormal shutdown of the shut-off valve, oil pump shutdown, and sudden increase in downstream distribution volume are set as targets for detection and identification. The simulation data of the software is used as the training data and test data of the embodiments of this specification, of which the training data accounts for 70% and the test data accounts for 30%. The accuracy of identifying abnormal causes reaches more than 95%; the accuracy of identifying abnormal causes reaches more than 90%. Some detection and identification results are shown as follows. Figure 2 and Figure 3 shown.

[0087] See also Figure 4 At the software level, the embodiments of this specification also provide a device for identifying the cause of abnormal operation of a transportation pipeline, which may specifically include the following structural modules.

[0088] An acquisition module 410 is used to acquire operation data of different positions of the transmission pipeline;

[0089] A clustering module 420 is used to cluster the operation data at different locations to obtain a preset number of operation data with different abnormal fluctuation types;

[0090] Establishing module 430, for establishing labels for the preset number of operation data with different abnormal fluctuation types according to the correspondence between the abnormal fluctuation type and the abnormal cause; the label is the abnormal cause corresponding to the abnormal fluctuation type in the operation data;

[0091] A calculation module 440 is used to use the operation data with labels as training samples and calculate the classification conditions of the operation data in the training samples;

[0092] The identification module 450 is used to identify the abnormal cause of the operation data to be identified according to the classification condition.

[0093] It should be noted that the various embodiments in this specification are described in a progressive manner, and the same or similar parts between the various embodiments can be referred to each other, and each embodiment focuses on the differences from other embodiments. In particular, for the device embodiment and the equipment embodiment, since they are basically similar to the method embodiment, the description is relatively simple, and the relevant parts can be referred to the partial description of the method embodiment.

[0094] After reading this specification, those skilled in the art can conceive of any combination of some or all of the embodiments listed in this specification without creative work, and these combinations are also within the scope of disclosure and protection of this specification.

[0095] In the 1990s, improvements to a technology could be clearly distinguished as hardware improvements (for example, improvements to the circuit structure of diodes, transistors, switches, etc.) or software improvements (improvements to the method flow). However, with the development of technology, many improvements to the method flow today can be regarded as direct improvements to the hardware circuit structure. Designers almost always obtain the corresponding hardware circuit structure by programming the improved method flow into the hardware circuit. Therefore, it cannot be said that an improvement in a method flow cannot be implemented using a hardware entity module. For example, a programmable logic device (PLD) (such as a field programmable gate array (FPGA)) is such an integrated circuit whose logical function is determined by the user's programming of the device. Designers can "integrate" a digital system on a PLD by programming it themselves, without having to ask a chip manufacturer to design and produce a dedicated integrated circuit chip. Moreover, nowadays, instead of manually making integrated circuit chips, this kind of programming is mostly implemented by "logic compiler" software, which is similar to the software compiler used when developing and writing programs, and the original code before compilation must also be written in a specific programming language, which is called hardware description language (HDL). There is not only one HDL, but many kinds, such as ABEL (Advanced Boolean Expression Language), AHDL (Altera Hardware Description Language), Confluence, CUPL (Cornell University Programming Language), HDCal, JHDL (Java Hardware Description Language), Lava, Lola, MyHDL, PALASM, RHDL (Ruby Hardware Description Language), etc. The most commonly used ones are VHDL (Very-High-Speed ​​Integrated Circuit Hardware Description Language) and Verilog2. Those skilled in the art should also know that it is only necessary to program the method flow slightly in the above-mentioned hardware description languages ​​and program it into the integrated circuit, and then it is easy to obtain the hardware circuit that implements the logic method flow.

[0096] The systems, devices, modules or units described in the above embodiments may be implemented by computer chips or entities, or by products with certain functions. A typical implementation device is a computer. Specifically, the computer may be, for example, a personal computer, a laptop computer, a cellular phone, a camera phone, a smart phone, a personal digital assistant, a media player, a navigation device, an email device, a game console, a tablet computer, a wearable device, or a combination of any of these devices.

[0097] It can be known from the above description of the implementation mode that the technicians in this field can clearly understand that the present specification can be implemented by means of software plus the necessary general hardware platform. Based on such an understanding, the technical solution of the present specification can be essentially or the part that contributes to the prior art can be embodied in the form of a software product, which can be stored in a storage medium, such as ROM / RAM, a disk, an optical disk, etc., including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment of the present specification or some parts of the embodiments.

[0098] Each embodiment in this specification is described in a progressive manner, and the same or similar parts between the embodiments can be referred to each other, and each embodiment focuses on the differences from other embodiments. In particular, for the system embodiment, since it is basically similar to the method embodiment, the description is relatively simple, and the relevant parts can be referred to the partial description of the method embodiment.

[0099] This specification can be used in many general or special computer system environments or configurations, such as personal computers, server computers, handheld or portable devices, tablet devices, multiprocessor systems, microprocessor-based systems, set-top boxes, programmable consumer electronic devices, network PCs, minicomputers, mainframe computers, and distributed computing environments that include any of the above systems or devices.

[0100] This specification may be described in the general context of computer-executable instructions executed by a computer, such as program modules. Generally, program modules include routines, programs, objects, components, data structures, etc. that perform specific tasks or implement specific abstract data types. This specification may also be practiced in distributed computing environments where tasks are performed by remote processing devices connected through a communication network. In a distributed computing environment, program modules may be located in local and remote computer storage media, including storage devices.

[0101] Although the present specification is described through embodiments, those skilled in the art will appreciate that there are many modifications and changes to the present specification without departing from the spirit of the present specification, and it is intended that the appended claims include these modifications and changes without departing from the spirit of the present specification.

Claims

1. A method for identifying the cause of abnormal operation of a pipeline, characterized in that: The method comprises: Obtain operating data at different locations of the transmission pipeline; Clustering the operation data at different positions to obtain a preset number of operation data with different abnormal fluctuation types; including: reconstructing the operation data in phase space, mapping the operation data into vectors in phase space to obtain phase space reconstructed data, clustering the phase space reconstructed data using a K-means clustering algorithm to obtain a preset number of operation data with different abnormal fluctuation types, wherein the preset number is the number of abnormal fluctuation types of the transmission pipeline; According to the correspondence between the abnormal fluctuation type and the abnormal cause, a label is established for the preset number of operation data with different abnormal fluctuation types; the label is the abnormal cause corresponding to the abnormal fluctuation type in the operation data; Using the running data with labels as training samples, calculating the classification conditions of the running data in the training samples; Extract the eigenvalue vector of the operating data to be identified, substitute the eigenvalue vector of the operating data to be identified into the classification condition, obtain the classification value of the eigenvalue vector of the operating data to be identified, and determine the abnormal cause corresponding to the classification value of the eigenvalue vector of the operating data to be identified according to the corresponding relationship between the classification value and the abnormal cause.

2. The method according to claim 1, characterized in that The operating data includes pressure data or flow data.

3. The method according to claim 1, characterized in that The method further comprises: Preprocessing the acquired operation data at different positions of the transmission pipeline; the preprocessing includes filling in missing data in the operation data by interpolation; Accordingly, the preprocessed operating data is clustered to obtain a preset number of operating data with different abnormal fluctuation types.

4. The method according to claim 3, characterized in that The filling of missing data in the operation data by interpolation includes: y1=y0α0(x1)+y2α1(x1)+y′0β0(x1)+y′2β1(x1) Among them, (x1, y1) is the missing data between the running data (x0, y0) and (x2, y2); x0, x1, x2 are the position coordinates; y0, y1, y2 are the values ​​of the running data; α and β are interpolation basis functions, which are obtained by the missing point coordinate x1 and the adjacent point coordinates x0 and x2.

5. The method according to claim 1, characterized in that The phase space reconstruction data includes: {y1, y2,..., y n }={from n ,...,from n+(d-1)τ } Among them, y1, y2, ..., y n is the value of the running data at different locations; n ,...,z n+(d-1)τ is the phase space vector after mapping, i.e., the phase space reconstruction data; d represents the dimension of the phase space; τ represents the delay time.

6. The method according to claim 1, characterized in that The calculation of the classification conditions of the running data in the training sample includes: Extracting the eigenvalue vector of the running data in the training sample; Based on the extracted eigenvalue vector of each running data in the training sample, a machine learning algorithm is used to calculate the classification conditions of the running data in the training sample.

7. The method according to claim 6, characterized in that The machine learning algorithm includes a logistic regression algorithm, a decision tree algorithm or a support vector machine algorithm.

8. A device for identifying the cause of abnormal operation of a pipeline, characterized in that: The device comprises: An acquisition module, used to acquire operation data of different positions of the transmission pipeline; A clustering module is used to cluster the operation data at different positions to obtain a preset number of operation data with different abnormal fluctuation types; the module includes: reconstructing the operation data in phase space, mapping the operation data into vectors in phase space, obtaining phase space reconstructed data, clustering the phase space reconstructed data using a K-means clustering algorithm, and obtaining a preset number of operation data with different abnormal fluctuation types, wherein the preset number is the number of abnormal fluctuation types of the transmission pipeline; An establishment module is used to establish a label for the preset number of operation data with different abnormal fluctuation types according to the corresponding relationship between the abnormal fluctuation type and the abnormal cause; the label is the abnormal cause corresponding to the abnormal fluctuation type appearing in the operation data; A calculation module, used to use the operation data with labels as training samples and calculate the classification conditions of the operation data in the training samples; The identification module is used to extract the eigenvalue vector of the operating data to be identified, substitute the eigenvalue vector of the operating data to be identified into the classification condition, obtain the classification value of the eigenvalue vector of the operating data to be identified, and determine the abnormal cause corresponding to the classification value of the eigenvalue vector of the operating data to be identified according to the corresponding relationship between the classification value and the abnormal cause.

Images