Pipeline leakage detection and leakage location positioning device and method

Abstract

The present invention discloses a pipeline leakage detection and leakage position positioning device and method, including: an inlet pressure sensor and an inlet flowmeter are installed at the inlet of the pipeline to be tested; an outlet pressure sensor and an outlet flowmeter are installed at the outlet of the pipeline to be tested; a node pressure sensor is installed at a preset node of the pipeline to be tested; the output ends of the inlet pressure sensor, the inlet flowmeter, the node pressure sensor, the outlet pressure sensor and the outlet flowmeter are all connected to the input end of the acquisition card, and the output end of the acquisition card is connected to the input end of the particle filter; wherein the number of node pressure sensors is one or more; the particle filter completes detection and positioning based on a particle filter algorithm. The present invention requires a small number of sensors, takes a short time to detect leakage, and has high positioning accuracy; it can detect and locate slow leakage and small leakage in long-distance oil pipelines; and has good real-time performance and large-scale application prospects.

Description

Technical Field

[0001] The present invention belongs to the technical field of pipeline leakage detection, and in particular relates to a pipeline leakage detection and leakage position positioning device and method. Background Art

[0002] Pipeline transportation is the main mode of transportation for oil and gas resources. With the large-scale application of pipeline transportation, the frequency of pipeline leakage accidents caused by natural reasons such as pipe aging, corrosion, water hammer process and human reasons such as illegal drilling and construction damage continues to increase. According to statistics, in 2016, my country's oil and gas pipelines reached 113,400 kilometers, of which the perforation rate of crude oil pipelines was about 0.66 times / km. The leakage of oil pipelines not only caused huge losses to the national economy, but also caused serious damage to the natural environment. In view of this, efficient and accurate status detection and leakage location of oil pipelines can obtain leakage information in time, reduce the economic losses and environmental damage caused by leakage accidents, and have important social significance and economic value.

[0003] Among the existing oil pipeline detection technologies, the in-pipe detection ball method is the main means of detecting short-distance pipelines; the acoustic detection method and the negative pressure wave detection method are the main means of detecting long-distance pipelines. The in-pipe detection ball method cannot perform real-time leak detection, and the detection takes a long time; the pressure wave detection method has poor detection effects on slow leaks and small leaks, and cannot detect leaks that have occurred; although the acoustic detection method is far superior to the negative pressure wave detection method in performance, it still has problems with sensitivity to slow leaks and small leaks and poor positioning accuracy. Although other oil pipeline detection technologies, such as the optical fiber method, magnetic leakage method, and robot detection method, have high efficiency and accuracy in locating pipeline leaks, they are difficult to use on a large scale due to their high economic costs and single applicable conditions.

[0004] In summary, a new type of oil pipeline leakage detection and positioning device is urgently needed. Summary of the invention

[0005] The purpose of the present invention is to provide a pipeline leakage detection and leakage location positioning device and method to solve one or more of the above-mentioned technical problems. The device of the present invention requires a small number of sensors, takes a short time to detect leakage, and has high positioning accuracy; it can detect and locate slow leakage and small leakage in long-distance oil pipelines; and it has good real-time performance and large-scale application prospects.

[0006] In order to achieve the above object, the present invention adopts the following technical solutions:

[0007] A pipeline leakage detection and leakage position positioning device comprises: an inlet pressure sensor, an inlet flow meter, a node pressure sensor, an outlet pressure sensor, an outlet flow meter, an acquisition card and a particle filter; the inlet pressure sensor and the inlet flow meter are installed at the inlet of the pipeline to be tested, and are used to detect the pressure value and flow rate of the medium at the inlet of the pipeline to be tested; the outlet pressure sensor and the outlet flow meter are installed at the outlet of the pipeline to be tested, and are used to detect the pressure value and flow rate of the medium at the outlet of the pipeline to be tested; the node pressure sensor is installed at a preset node of the pipeline to be tested, and is used to detect the pressure value of the medium at the preset node of the pipeline to be tested; the output ends of the inlet pressure sensor, the inlet flow meter, the node pressure sensor, the outlet pressure sensor and the outlet flow meter are all connected to the input end of the acquisition card, and the output end of the acquisition card is connected to the input end of the particle filter; wherein the number of the node pressure sensors is one or more.

[0008] Furthermore, it also includes: a first auxiliary pipeline and a second auxiliary pipeline; the inlet of the pipeline to be tested is connected with the outlet oil storage tank through the first auxiliary pipeline, and the outlet of the pipeline to be tested is connected with the inlet oil tank through the second auxiliary pipeline; on the first auxiliary pipeline, a first ball valve, a pump, a check valve, a second ball valve and a control valve are sequentially arranged along the medium conveying direction; on the second auxiliary pipeline, a third ball valve is arranged; the first auxiliary pipeline and the second auxiliary pipeline have the same specifications as the pipeline to be tested.

[0009] Furthermore, the particle filter includes: a wireless transmission module, a data storage module, a computing module and a display module; the receiving end of the wireless transmission module is used to receive data transmitted by the acquisition card, the output end of the wireless transmission module is connected to the receiving end of the data storage module, the output end of the data storage module is connected to the input end of the computing module, and the output end of the computing module is connected to the input end of the display module; the computing module is a single-chip microcomputer with a preset computing capacity, and a preset particle filter algorithm program is integrated in the single-chip microcomputer.

[0010] A pipeline leak detection and leak location positioning method, comprising the following steps:

[0011] S1, collecting and acquiring basic data and working data of the pipeline to be tested; the working data at least includes the inlet pressure value and flow value of the pipeline to be tested, the outlet pressure value and flow value, and the pressure value at a preset node in the pipeline to be tested;

[0012] S2, according to the data collected in step S1, using a transient model method based on adding leakage information to establish a state space equation of the pipeline fluid;

[0013] S3, judging whether leakage occurs by the flow value collected in step S1; if leakage occurs, outputting the preliminary leakage amount and jumping to step S4; if leakage does not occur, jumping to step S1;

[0014] S4, according to the state space equation obtained in step S2, the particle filter and the preliminary leakage amount are initialized, and the pipeline state at the time of leakage is preliminarily estimated based on the particle filter algorithm, that is, all preset nodes are traversed in a preset order, and a set of estimated values ​​whose difference between the estimated value and the observed value is less than a preset threshold value is found, and the leakage amount, leakage coefficient and preliminary leakage position of the pipeline to be tested are obtained;

[0015] S5, by increasing the number of preset nodes, a state space equation that is more accurate than step S2 is established, and then the particle filter is initialized. The leakage position and leakage amount in the initialization condition are obtained by step S4, and the pipeline state at the time of leakage is estimated based on the particle filter algorithm. When the difference between the estimated value and the observed value is less than a preset threshold that is more accurate than step S4, the optimized leakage position is obtained.

[0016] A further improvement of the present invention is that step S1 specifically includes: collecting the length L of the pipeline to be measured, the inner diameter D of the pipeline to be measured, the hydraulic friction coefficient f, the pressure wave propagation speed C s , the horizontal angle of the pipeline θ; collect the pressure head value H1 of the pipeline inlet oil transfer station, the flow value Q1 and the pressure head value H1 of the outlet oil receiving station n , flow value Q n ; Collect the pressure head value H at the preset node position of the pipeline i′ ; Collect the density ρ of the transport medium in the pipeline.

[0017] A further improvement of the present invention is that step S2 specifically includes: establishing a mathematical model of the pipeline fluid through mass balance and momentum balance of the conveying medium, and using the characteristic line method to obtain a numerical solution thereof;

[0018] The expression of the pipeline state space equation including leakage is constructed as follows:

[0019] x j+1 =Ax j +Bu j +Du j 2 +Eu j+1 +aξ+FQ l ;

[0020] y=Cx j

[0021] In the formula, the state of the pipe flow system is x = [H1, ..., H n-1 , Q2, ..., Q n ] T , j represents the time step of the pipeline system, n is the number of nodes in the pipeline, u j is the control quantity of the system, λ l is the leakage coefficient of the leakage point; H lis the pressure head value at the leak point, y measurement value, and the output matrix C is: Q l is the leakage measured by the flow meter, Q l =Q n -Q1.

[0022] A further improvement of the present invention is that step S3 specifically includes:

[0023] Set the threshold th1 and the response delay t; when |Q1-Q n |>th1, and this state lasts for Δt>t, then it is judged that the pipeline is leaking; otherwise, it is judged that the pipeline is not leaking; when a leak occurs, the leakage time t′ is output, and the average leakage amount within Δt is output. As a replacement for the pipeline state space equation Q l Use; where Q1 is the flow value detected at the pipeline inlet, Q n The flow rate value detected at the pipe outlet.

[0024] A further improvement of the present invention is that step S4 specifically includes:

[0025] S4.1: Initialize the parameters in the particle filter algorithm: Initialize the matrix F, add the mean leakage value along the pipeline direction in the n-2 node positions where the pipeline may leak

[0026] Initialize the state space equation, set time j = 0, leakage time j = t'; input the working state of the pipeline transport medium as the control quantity u of the state space equation j ,u j =[Q 1，j , H n，j ] T ; Set the number of particles to P, initialize the pipeline state particle set χ and the leakage particle set τ; where the mean is Generate the leakage particle set τ using the normal distribution of;

[0027] S4.2: The initial state x of the pipeline from step S4.1 j , control quantity u j ,u j+1 Calculate the pipeline status x j+1 Based on the particle set x in step S4.1 j , τ j And the observed value y=[H 1，j , Q n，j ] T , calculate and obtain the particle set χ j+1 ;

[0028] Calculate x at time j j+1 With χ j+1The difference dis(i) is calculated as follows:

[0029] dis(i)=x j+1 -x j+1

[0030] Calculate the particle set χ j+1 The weights w(i) corresponding to the P particles in the equation are normalized to q(i). The calculation formula is:

[0031]

[0032]

[0033] Among them, i is the calculation step required for importance resampling, i∈[1,P]

[0034] By particle set χ j+1 and weight value q(i), we get a set of weighted particle sets The posterior density function that approximately represents the pipeline state at time j+1, where the weights satisfy the normalization condition

[0035] S4.3: For the weighted particle set χ′ obtained in step S4.2 j+1 Using the importance resampling method, χ′ j+1 After the particles with lower weights are replaced by particles with higher weights, they are used as the particle set to estimate the state of the pipeline at time j+2; before the particle filter estimation at time j+2, the particle set χ′ j+1 Update the updated particle set τ j Perform pre-processing;

[0036] S4.4: Repeat steps S4.2 and S4.3 to obtain the particle filter estimation value of the pipeline system at any time of leakage state Particle filter estimate of leakage

[0037] S4.5: Compare measured values ​​with estimated values ​​at each node in the piping system Compare the measured leakage Q l With estimated value In step S4.1, determine whether the leakage information attached to the node n along the pipeline is correct; if it is correct, the attachment of the leakage information is valid, and the estimated leakage position is output. If it is incorrect, update the leakage position n=n+1 in step S4.1 and repeat steps S4.2 to S4.5.

[0038] A further improvement of the present invention is that step S5 specifically includes:

[0039] Step S5.1: increasing the number of segments N′ of the pipeline to be tested in step S1 to establish a more accurate state space equation of the pipeline to be tested;

[0040] Step S5.2: Set a threshold value th2 that is more accurate than that in step S4. The additional leakage position and leakage amount in the initialization condition of the particle filter are the values ​​obtained in step S4. The leakage position The update method is

[0041] Where Δx′ i is the preliminary leak position in step S4 Δx′ is the distance between two nodes of the state space equation in step S5.1, and the value range of m is To round up to an integer;

[0042] Iterate and update until the preset threshold condition is met, and output accurate leakage location information Leakage coefficient λ of the leak hole l .

[0043] A further improvement of the present invention is that step S4.3 comprises:

[0044] (1) Generate a constant in the interval [0, 1] based on the normal distribution, denoted as r(i);

[0045] (2) The weight of the pipeline state at this moment obtained from step 4.2 Compare with r(i); if Update k = k + 1 to get Continue to compare with r(i); if The particle with the higher weight corresponding to weight q(i=k) Replace particles with lower weights

[0046] (3) According to the number of particles P in the particle filter, repeat steps (1) and (2) for P times; and Take the mean value to obtain the particle filter estimate of the pipeline state at time j+1 Particle filter estimation of leakage at time j+1

[0047] (4) Update particle set χ j+1 = x" j+1 , update the particle set τ j+1 =τ′ j+1 ;

[0048] The specific preprocessing method is: τ′j+1 (i) = τ′ j+1 (i)+δ(i), where δ(i) obeys a normal distribution with a mean of 0, and the required variance of the normal distribution depends on the actual leakage amount.

[0049] Compared with the prior art, the present invention has the following beneficial effects:

[0050] The device of the present invention requires fewer sensors, takes a short time to detect leaks, and has high positioning accuracy; it can detect and locate slow leaks and small leaks in long-distance oil pipelines; and it has good real-time performance and large-scale application prospects.

[0051] The particle filter-based oil pipeline leakage detection and positioning method of the present invention can realize the accurate positioning of the leakage position of the oil pipeline under the conditions of severe nonlinearity and non-Gaussian distribution of measurement noise; and while improving the positioning accuracy, it ensures that the amount of calculation does not increase significantly; the proposed leakage preprocessing method in particle filter estimation can effectively estimate the leakage under conditions such as non-constant pipeline leakage and slow leakage. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] Figure 1 It is a structural schematic diagram of an oil pipeline leakage detection and positioning device based on a particle filtering algorithm according to an embodiment of the present invention;

[0053] Figure 2 A schematic diagram of the arrangement position of pipeline sensors in an embodiment of the present invention;

[0054] Figure 3 A schematic flow chart of a method for detecting and locating oil pipeline leakage based on a particle filtering algorithm according to an embodiment of the present invention;

[0055] Figure 4 It is a flow chart of importance resampling and leakage preprocessing algorithm in particle filtering in an embodiment of the present invention;

[0056] Figure 5 It is a schematic diagram of the estimated value of the pressure head along the pipeline in the oil pipeline leakage detection method based on particle filtering according to an embodiment of the present invention; Figure 5 (a) is a schematic diagram showing the comparison between the measured value of the first node pressure sensor and the estimated value of the particle filter; Figure 5 (b) is a schematic diagram comparing the measured value of the second node pressure sensor and the estimated value of the particle filter; Figure 5 (c) is a schematic diagram comparing the measured value of the pressure sensor at the third node with the estimated value of the particle filter;

[0057] Figure 6 It is a schematic diagram of estimating pipeline leakage amount and leakage coefficient in the oil pipeline leakage detection method based on particle filtering in an embodiment of the present invention; Figure 6(a) is a schematic diagram of leakage estimation; Figure 6 (b) is a schematic diagram of leakage coefficient estimation;

[0058] In the figure, the outbound oil storage tank 1; the first ball valve 2; the pump 3; the check valve 4; the second ball valve 5; the control valve 6; the first auxiliary pipeline 7; the pipeline to be tested 8; the second auxiliary pipeline 9; the third ball valve 10; the inbound oil tank 11; the inlet pressure sensor 12; the inlet flow meter 13; the outlet pressure sensor 15; the outlet flow meter 16; the acquisition card 17; the particle filter 18; the first node pressure sensor 19; the second node pressure sensor 20; the third node pressure sensor 21. DETAILED DESCRIPTION

[0059] The following is a complete and clear description of the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings of the embodiments of the present invention. It should be understood that the embodiments described herein are only used to illustrate and explain the present invention and are not intended to limit the present invention. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.

[0060] See also Figure 1 and Figure 2 The present invention discloses a pipeline leakage detection and positioning device based on particle filtering, comprising: an outbound oil storage tank 1, a first auxiliary pipeline 7, a pipeline to be tested 8, a second auxiliary pipeline 9 and an inbound oil tank 11 which are connected in sequence.

[0061] The first auxiliary pipeline 7 is provided with a first ball valve 2, a pump 3, a first check valve, a second ball valve 5, a flow control valve and a second check valve in sequence along the conveying direction of the medium; the second auxiliary pipeline 9 is provided with a third ball valve 10 and a pressure control valve.

[0062] The first auxiliary pipeline 7 and the pipeline to be tested 8 are connected through a first connecting flange, and an inlet pressure sensor 12 and an inlet flow meter 13 are installed at the first connecting flange to detect the pressure value and flow rate of the medium at the inlet of the pipeline to be tested.

[0063] The second auxiliary pipeline 9 and the pipeline to be tested 8 are connected via a second connecting flange, and an outlet pressure sensor 15 and an outlet flow meter 16 are installed at the second connecting flange to detect the pressure value and flow rate of the medium at the outlet of the pipeline to be tested.

[0064] A plurality of preset nodes, which may be three, are arranged along the pipeline to be tested 8; node pressure sensors with equal precision are installed at the preset nodes to detect the pressure value of the medium at the preset nodes of the pipeline to be tested.

[0065] The data output ends of the inlet pressure sensor 12, the inlet flow meter 13, the node pressure sensor, the outlet pressure sensor 15 and the outlet flow meter 16 are all connected to the data input end of the acquisition card 17, and the output end of the acquisition card 17 is connected to the data input end of the particle filter 18. The particle filter 18 processes the input data and outputs an estimated predicted leakage position to complete the detection.

[0066] The acquisition card 17 includes a wired transmission and a wireless transmission module, and should be set near the pressure sensor along the pipeline to be tested 8. The measurement values ​​of the pressure sensors along the pipeline to be tested are collected by wired connection, and connected to the particle filter data storage module by wireless transmission, so as to realize the connection between the measurement values ​​of each sensor along the pipeline to be tested 8 and the particle filter data storage module.

[0067] Particle filter: The particle filter 18 includes a wireless transmission module, a data storage module, a calculation module and a display module. The wireless transmission module has the function of wirelessly transmitting information between the acquisition card and the particle filter; the data storage module is used to store the real-time working data of the inlet pressure sensor 12, the inlet flow meter 13, the node pressure sensor, the outlet pressure sensor 15 and the outlet flow meter 16 transmitted by the acquisition card; the calculation module integrates the preset particle filter algorithm and the state space equation of the pipeline to be tested, and has the calculation function to calculate the real-time working data stored in the data storage module; when the calculation module completes the detection, the display module is used to display the suspected leakage location.

[0068] The working principle of the device of the present invention:

[0069] When collecting the pressure head value H1 at the end of the first auxiliary pipeline, the pump, the first check valve, the flow control valve, and the pressure control valve are used to ensure that the pressure head value and flow value of the medium transported by the pipeline to be tested are in a steady state, and the steady-state flow value Q1 is stored in the data storage module of the particle filter as the control quantity u required for the particle filter. j The flow control valve and pressure control valve ensure that the flow value Q1 of the conveying medium remains unchanged under the leakage condition of the pipeline to be tested. When the collection conditions of the flow value Q1 are met, the pressure head value H at the first end of the second auxiliary pipeline measured by the outlet pressure sensor is n The data storage module stored in the particle filter, the pressure control valve is used to limit the maximum pressure head value of the conveying medium in the second auxiliary pipeline to not more than H n At the same time, the pressure head value and flow value are in a steady state, and the steady-state pressure head value H at the first end of the second auxiliary pipeline n As the control quantity u required for particle filtering j When the pipeline to be tested leaks, due to the influence of the leakage and the working requirements of the flow control valve 6 and the pressure control valve 13 of the present invention, the measured value of the pressure head value H1 and the flow value Qn Will drop to a new steady state, in order to ensure the flow value Q1 and pressure head value H of the pipeline to be tested under leakage conditions n Compared with the flow value Q1 and pressure head value H of the pipeline to be tested under normal working conditions n The delivery pressure provided by the pump should be at least 5% to 10% greater than the delivery pressure required under actual working conditions. The check valve prevents leakage from causing transient changes in the pipeline to be tested, which causes a small reflux of the conveying medium, thereby causing the flow value of the first auxiliary pipeline to fluctuate. Observation values ​​H1, Q n The data is transmitted to the storage module of the particle filter in real time. The pressure head value measurement value at the middle position of the pipeline to be tested is collected through the node pressure sensor along the pipeline to be tested, and stored in the data storage module of the particle filter.

[0070] To ensure the effectiveness of the device and method of the present invention, when the data collection conditions are met, the operating conditions of various control elements such as the speed of the pump, the flow control valve, and the pressure control valve should remain constant. If the actual pipeline causes the pressure value or flow value of the conveying medium to change due to any reason, the debugging of the above control elements should be repeated, and the particle filter should be initialized. After the above data collection conditions are met again, the collected data should be re-entered into the storage module of the particle filter.

[0071] The working process of the device of the present invention:

[0072] In the device of the present invention, along the medium conveying direction, the outbound oil storage tank 1 is connected and communicated with the first ball valve 2 and the pump 3 in sequence, the pump 3 is connected and communicated with the check valve 4, the second ball valve 5 and the control valve 6 in series, the control valve 6 is connected and communicated with the first auxiliary pipeline 7 and the pipeline to be tested 8 in sequence, and the pipeline to be tested 8 is connected and communicated with the second auxiliary pipeline 9, the third ball valve 10 and the inbound oil tank 11 in sequence. The pump 3, the check valve 4 and the control valve 6 control the starting flow and the end pressure of the pipeline to be tested 8 required for the particle filter. The inlet pressure sensor 12 and the inlet flow meter 13 are arranged at the end of the first auxiliary pipeline 7, and the outlet pressure sensor 15 and the outlet flow meter 16 are arranged at the head end of the second auxiliary pipeline 9, which are used to collect the more accurate measurement values ​​required by the particle filter 18. The specifications of the first auxiliary pipeline 7 and the second auxiliary pipeline 9 are the same as those of the pipeline to be tested 8, so as to further ensure the accuracy of the measurement values; a number of node pressure sensors are arranged at preset node positions along the pipeline to be tested 8, which are used to identify the validity of the pipeline state estimation value obtained by the particle filter 18.

[0073] The specific node setting method includes: before arranging the node pressure sensor, the pipeline 8 to be tested must first be segmented, that is, Wherein, [n] is the integer of the node after the pipeline is segmented, the number of pipeline nodes n=N+1, Δx is the distance between two pipeline nodes, and L is the total length of the pipeline to be tested. The layout positions of several node pressure sensors should be on the preset nodes of the pipeline to be tested 8. The sensors along the pipeline to be tested 8 are connected to the particle filter through an acquisition card. The acquisition card has an input module, an output module and a data transmission module, which can realize the real-time data transmission function required by the particle filter.

[0074] See also Figure 2 , which shows a specific arrangement of sensors for the oil pipeline to be tested. It should be noted that Figure 2 Only the pipeline of the specific embodiment is specifically illustrated, and the number of nodes, the distance between adjacent nodes, etc. shown therein can be set and modified according to actual needs, and are not intended to be a specific limitation of the technical solution of the present invention.

[0075] See also Figure 3 The present invention provides a method for detecting and locating oil pipeline leakage based on a particle filtering algorithm, which specifically comprises the following steps:

[0076] Step 1: Collect the parameters of the pipeline to be tested, collect the pressure value and flow value data at the beginning and end of the pipeline to be tested, collect the internal pressure data of the pipeline at the middle node position of the pipeline to be tested, and collect the parameters of the transport medium in the pipeline to be tested.

[0077] Step 1 specifically includes: collecting the length L of the pipeline to be tested, the inner diameter D of the pipeline to be tested, the hydraulic friction coefficient f, and the pressure wave propagation speed C s , the horizontal angle of the pipeline θ; collect the pressure head value H1 of the pipeline inlet oil transfer station, the flow value Q1 and the pressure head value H1 of the outlet oil receiving station n , flow value Q n ; Collect the pressure head value H at the middle position of the pipeline i′ ; Collect the density ρ of the transport medium in the pipeline.

[0078] Step 2: According to the parameters of step 1, a transient model method based on adding leakage information is used to establish the state space equation of the pipeline fluid.

[0079] Step 2 specifically includes: establishing a mathematical model of the pipeline fluid through mass balance and momentum balance of the conveying medium, and using the characteristic line method to calculate the solution value.

[0080] The state space equation of the pipe flow system can be expressed as equation (1) and equation (2):

[0081] x j+1 =ψ(x j ,u j ,u j+1 ) (1)

[0082] y=Cx j(2)

[0083] Specifically:

[0084] x j+1 =Ax j +Bu j +Du j 2 +Eu j+1 +aξ (3)

[0085] y=Cx j (4)

[0086] In the above formula, the state of the pipe flow system is x = [H1, ..., H n-1 , Q2, ..., Q n ] T , j represents the time step of the pipeline system, u j is the control quantity of the system.

[0087] Furthermore, the leakage model is added Obtain the pipeline state equation including the leak:

[0088] x j+1 =Ax j +Bu j +Du j 2 +Eu j+1 +aξ+FQ l (5)

[0089] y=Cx j (6)

[0090] In the formula, λ l is the leakage coefficient of the leakage point; H l is the pressure head value at the leakage point.

[0091] Step 3: Based on the flow measurement value Q1 at the head end of the pipeline to be tested and the flow measurement value Q at the end of the pipeline to be tested n Determine whether there is a leak in the pipeline.

[0092] Step 3 specifically includes: setting the threshold Q L , set the response delay t. When |Q1-Q n |>Q L , and this state lasts for Δt′>t, then it can be determined that the pipeline has leaked. Output the leaking time t′, and output the mean value of the leakage amount within Δt′ To replace the measured leakage Q in equation (5) l .

[0093] Step 4: Initialize the particle filter and leakage amount, make a preliminary estimate of the pipeline state at the time of leakage based on the particle filter algorithm, and then obtain the pipeline leakage amount, leakage coefficient and leakage location.

[0094] Step 4 specifically includes:

[0095] Step 4.1: Initialize the parameters in the particle filter algorithm. Initialize the matrix F, add the mean value of the leakage along the pipeline direction in the n-2 locations where the pipeline may leak At this time, the state space equation in the particle filter is equation (5). Initialize equation (5) and set time j = 0 and leakage time j = t'. Input the working state of the pipeline transport medium as the control quantity u of the state space equation j , specifically u j =[Q 1，j , H n，j ] T . Set the number of particles to P and initialize the pipeline state particle set χ j and the leakage particle set τ j Specifically, with the mean value u j＝0 The normal distribution generates the pipeline state particle set χ j , with the mean value being The normal distribution of the leakage particle set τ j , generate the leakage particle set τ j The required variance should satisfy the fluctuation range not exceeding 20% of.

[0096] Step 4.2: The initial state x of the pipeline in step 4.1 j , control quantity u j ,u j+1 Calculate pipeline status x j+1 Based on the particle set x in step 4.1 j , τ j , and the observed value y=[H 1，j , Q n，j ] T , calculate the particle set χ j+1 ; Calculate at this moment, x j+1 With χ j+1 The difference dis(i):

[0097] dis(i)=x j+1 -x j+1 (7)

[0098] According to formula (8), calculate the particle set χ j+1 The weights w(i) corresponding to the P particles in the equation are normalized to q(i):

[0099]

[0100]

[0101] Where i is the calculation step required for importance resampling, i∈[1,P]. j+1 and weight value q(i), we get a set of weighted particle sets The posterior density function that approximately represents the pipeline state at time j+1, where the weights satisfy the normalization condition

[0102] Step 4.3: For the weighted particle set χ′ in step 4.2 j+1 Using the importance resampling method, χ′ j+1 After the particles with lower weights are replaced by particles with higher weights, they are used as the particle set to estimate the state of the pipeline at time j+2. Before the particle filter estimation at time j+2, j+1 Update the updated particle set τ j Perform preprocessing.

[0103] Repeat steps 4.2 to 4.3 to obtain the particle filter estimation value of the pipeline system at any time of leakage state. Particle filter estimate of leakage

[0104] Step 4.4: Compare the measured values ​​with the estimated values ​​at each node in the piping system Compare the measured leakage Q l With estimated value Determine whether the leakage information attached to the node n along the pipeline in step 4.1 is correct. If so, the attachment of the leakage information is valid, and output the estimated leakage location Then proceed to step 5 to further estimate the leakage position; if not, update the leakage position n=n+1 of step 4.1, and repeat steps 4.2 to 4.4.

[0105] Step 5: Input the preliminary estimated pipeline leakage amount, use the particle filter algorithm to accurately estimate the pipeline state at the time of leakage, and output the accurate leakage location information.

[0106] Step 5 specifically includes:

[0107] Step 5.1: According to equation (5), establish a more accurate state space equation of the pipeline. The specific implementation method is to obtain the pipeline leakage position according to step 4. By increasing the number of computational nodes n in the state space equation of the pipeline, the distance between the actual position Z of the leakage point and the computational node is made as close as possible;

[0108] Step 5.2: The core algorithm and specific implementation of this step are the same as those of steps 4.1 to 4.3. The particle filter initialization condition and threshold th i According to this step, the accurate information of the pipeline leakage location can be obtained.

[0109] The present invention proposes a method for detecting and locating oil pipeline leakage based on particle filtering, which can accurately locate the leakage position of the oil pipeline under the conditions of severe nonlinearity and non-Gaussian distribution of measurement noise; and while improving the positioning accuracy, it ensures that the amount of calculation does not increase significantly; the proposed preprocessing method of leakage amount in particle filtering estimation can effectively estimate the leakage amount of the pipeline under conditions such as non-constant leakage amount and slow leakage.

[0110] Example

[0111] The oil pipeline leakage detection and positioning method based on particle filtering in the particle filter of the present invention comprises the following steps:

[0112] Step 1: Collect basic data of the pipeline in the embodiment. Specifically, the length L, inner diameter D, hydraulic friction coefficient f, pressure wave propagation speed C of the pipeline in the embodiment s , the horizontal angle of the pipeline θ; collect the density ρ of the medium transported by the pipeline, and determine the number of pipeline sections N′ in the embodiment.

[0113] Collect the working data of the pipeline in the embodiment. Specifically, it includes: collecting the pressure head value H1 at the end of the auxiliary pipeline, the flow value Q1, the pressure head value H1 at the head end of the auxiliary pipeline n , Q n ; Collect the pressure head value H along the pipeline i′ .

[0114] Step 2: Establish a mathematical model of pipeline fluid through mass and momentum balance of the transported medium. This mathematical model can be represented by a set of hyperbolic partial differential methods. Use the finite difference method based on the characteristic line method to solve its numerical solution. Its stability condition is Where Δt is the calculation step size.

[0115] The state space equations of the pipe flow system are shown in equations (10) and (11):

[0116] x j+1 =ψ(x j ,u j ,u j+1 ) (10)

[0117] y=Cx j (11)

[0118] Specifically:

[0119] x j+1=Ax j +Bu j +Du j 2 +Eu j+1 +aξ (12)

[0120] y=Cx j (13)

[0121] In the above formula, the state of the pipe flow system is x = [H1, ..., H n-1 , Q2, ..., Q n ] T , j represents the time step of the pipeline system, n is the number of nodes N in the pipeline, and the control quantity u j =[Q 1，j , H n，j ] T ; Observed value y = [H 1，j , Q n，j ] T .

[0122] Among them, the matrix A is:

[0123]

[0124] The matrix B is:

[0125] The matrix D is:

[0126] The matrix E is:

[0127] The nonlinear term ξ is:

[0128] The coefficient matrix a is:

[0129]

[0130] The output matrix C is:

[0131] Furthermore, the leakage information equation Add equations (12) and (13):

[0132] x j+1 =Ax j +Bu j +Du j 2 +Eu j+1 +aξ+FQ l (14)

[0133] y=Cx j (15)

[0134] Among them, λ l is the leakage coefficient of the leakage point; H l is the pressure head value at the leakage point.

[0135] Step 3: Set the threshold QL, set the response delay t, and the abnormal state duration Δt′. Introduce the judgment condition: |Q1-Q n |>Q L , Δt′>t, then it can be judged that the pipeline has leaked, and the leakage time t′ is output, and the mean value of the leakage amount within Δt is output As the initialization parameter in step 4. Otherwise, it is determined that the pipeline has not leaked, and step 3 is repeated.

[0136] Due to the influence of noise, the specific leakage amount cannot be calculated based on |Q1-Q n | It is concluded that the mean value of the leakage amount within the output Δt′ is Used as the particle filter initialization parameter in step 4, where i″ is the mean leakage value The sampling start time, Δt′ is the time when the pipeline state returns to the steady state after the leakage, i″∈[1,Δt′].

[0137] The step 4 specifically includes:

[0138] Step 4.1: Initialize the particle filter. Initialize the matrix F = [1 0 … 0 0] so that the leakage is added to the node n = 2 of the pipeline to be tested, and add the leakage mean to the nodes along the pipeline. Initialize the time j = 0 in equation (5), and initialize the leakage time j = t′. Set the number of particles to P, initialize the pipeline state particle set χ, and initialize the leakage particle set τ. Specifically, take the mean value u j＝0 The particle set χ generated by the normal distribution j , with the mean value being The particle set τ is generated by the normal distribution j .

[0139] It should be noted that this step and other steps including particle filter estimation specifically estimate the state of the pipeline using the control value u and the observation value y. The remaining measurement values ​​that can be obtained are only used to identify the correctness of the particle filter's estimation of the pipeline state, and are not used as specific parameters in the particle filter estimation.

[0140] Step 4.2: x obtained from step 4.1 j ,u j ,u j+1 , calculate the pipeline state x j+1 ; The particle set x obtained by step 4.1 j , τ j, observed value y=[H 1，j , Q n，j ] T , calculate the particle set χ j+1 ; and calculate at this moment, x j+1 With χ j+1 The difference dis(i):

[0141] dis(i)=x j+1 -x j+1 (16)

[0142] According to formula (16), calculate the particle set χ j+1 The weights w(i) corresponding to the 100 particles in the image are normalized to q(i):

[0143]

[0144]

[0145] Among them, i=1, 2,...,100, μ=0, σ=1.

[0146] By particle set χ j+1 and weight value q(i), we can get a set of weighted particle sets To approximate the posterior density function of the pipeline state at time j+1, where the weights satisfy the normalization condition

[0147] Step 4.3: The weighted particle set χ′ in step 4.2 j+1 Using the importance resampling method, χ′ j+1 The particles with lower weights are replaced by particles with higher weights, which are used as the particle set to estimate the state of the pipeline at time j+2. Before the particle filter estimation at time j+2, j+1 Update the updated particle set τ j Preprocessing is performed. The flowchart of the specific algorithm is as follows Figure 4 shown.

[0148] The specific implementation steps of this step are:

[0149] (1) Generate a constant in the interval [0, 1] based on the normal distribution, denoted as r(i);

[0150] (2) The weight of the pipeline state at this moment obtained by step 4.2 Compare with r. If Update k = k + 1 to get Continue to compare with r. The particle with the higher weight corresponding to weight q(i=k) Replace particles with lower weights

[0151] (3) According to the number of particles in the particle filter P = 100, repeat steps (1) and (2) 100 times. and Taking the mean, we can get the particle filter estimate of the pipeline state at time j+1 Particle filter estimation of leakage at time j+1

[0152] (4) Update particle set χ j+1 = x" j+1 , update the particle set τ j+1 =τ′ j+1 To overcome the decrease in particle diversity in the leakage, the leakage coefficient λ l There may be non-constant problems, and when updating the particle set τ j+1 Before, for the particle set τ′ j+1 Preprocessing is performed. The specific preprocessing method is: τ′ j+1 (i) = τ′ j+1 (i)+δ(i), where δ(i) follows a normal distribution with mean 0.

[0153] Repeat steps 4.2 to 4.3 to obtain the particle filter estimate of the pipeline state of the embodiment at any time. Particle filter estimate of leakage

[0154] Step 4.4: Compare the measured values ​​at three locations along the pipeline of the embodiment (the measured values ​​of the first node pressure sensor 19, the second node pressure sensor 20, and the third node pressure sensor 21) with the estimated values Comparative leakage Q l With estimated value Determine whether the leak location added in step 4.1 is correct. Note that the estimated value obtained in this step is The position is related to the number of segments N' of the example pipeline in step 1. If the estimated value The location is not located on the node in step 1, and the estimated value can be Make appropriate compensation, and the specific compensation value is based on the measured value in steady state.

[0155] The specific implementation steps of this step are:

[0156] (1) According to the embodiment, the three node measurement values ​​along the pipeline are set to an equal threshold value th1. l Set a threshold th0 = 5 × 10 -4 m 3 / s;

[0157] (2) Determine whether the difference between the measured values ​​of the three nodes along the pipeline of the embodiment and the estimated values ​​of the three nodes is less than the threshold value th1, and whether the difference between the leakage amount and the estimated leakage amount is less than the threshold value th0. If so, the addition of leakage information is valid, and the estimated leakage location is output. Then proceed to step 5 to further estimate the leakage position; if not, update the leakage position n=n+1 of step 4.1, and repeat steps 4.2 to 4.4.

[0158] The step 5 specifically includes:

[0159] Step 5.1: According to formula (14), a more accurate state space equation of the pipeline is established. The specific implementation method is to increase the number of segments N' of the pipeline in the embodiment in step 1. The specific increase in the number of segments N' depends on the actual working conditions, but in principle, the number of segments N in this step should be at least twice the number of segments N' in step 1, that is, N ≥ 2N';

[0160] Step 5.2: The core algorithm and specific implementation of this step are the same as those of steps 4.1 to 4.3, except that the particle filter initialization condition and threshold th1 are different. The specific differences are:

[0161] (1) The additional leakage position and leakage amount in the initialization condition of the particle filter in this step are given by step 4. The leakage position The update method is Where Δx′ in The leak location in step 4 Δx′ is the distance between the two nodes of the state space equation in step 5.1, and the value range of m is To round up;

[0162] (2) The new threshold th′1 should satisfy th′1<th1, th′0<th0, so as to achieve the requirement of accurately locating the leakage point; specifically, when the difference between the measured values ​​of the three nodes along the pipeline of the embodiment and the estimated values ​​of the three nodes is less than the threshold th′1, the difference between the leakage amount and the estimated leakage th0 outputs the accurate leakage location information (Specifically, the distance from the oil delivery station), the leakage coefficient λ of the leakage hole l ; Otherwise repeat step 5.2.

[0163] The significance of this step is that it solves the problem that the amount of particle filtering calculation increases sharply when the accuracy of the particle filter for the pipeline leakage location in the embodiment is improved. According to the number of nodes n of the pre-segmented pipeline in the embodiment, the number of times the particle filter algorithm estimates the leakage point location in step 4 is at least 1 time and at most n-1 times. The uncertainty of the actual leakage location is the main reason for the increase in the amount of particle filtering algorithm calculation. This step is performed by inputting the preliminary estimate of the leakage location output in step 4. Propose a new way to update the leak location The advantage of this new method of updating the leakage position is that, even if the number of computational nodes of the state-space equation re-established in step 5.1 increases exponentially, the particle filter algorithm updates the leakage position estimate at most m times, thereby achieving the purpose of reducing the computational complexity of the particle filter algorithm while improving the accuracy of the leakage position estimation.

[0164] See also Figure 5 and Figure 6 The actual effect of this method can be seen from the figure: the embodiment pipeline with a length of 37 km leaked at about 11.1 km from the oil inlet station. The transportation volume of the embodiment pipeline is Q1 = 0.081m 3 / s, the leakage is 0.0028m 3 / s, which is about 3.5% of the transport capacity. The leakage coefficient of the pipeline leakage point in the embodiment is about 1.4×10 -4 The error between the leakage position of the pipeline in the embodiment and the leakage position estimated by the filter is about 141m, and the positioning error is about 3.8%. When the number of segments of the pipeline is increased, the positioning accuracy of the leakage point is higher. This result shows the effectiveness of this method.

[0165] The above-mentioned embodiment only expresses one implementation mode of the present invention, and its description is relatively specific and detailed. A person skilled in the art may modify the technical solution or some parameters thereof without departing from the principle of the present invention, and all of these belong to the protection scope of the present invention.

Claims

1. A method for locating a pipeline leakage detection and leakage position locating device, characterized in that: The pipeline leakage detection and leakage position positioning device comprises: an inlet pressure sensor (12), an inlet flow meter (13), a node pressure sensor, an outlet pressure sensor (15), an outlet flow meter (16), an acquisition card (17) and a particle filter (18); An inlet pressure sensor (12) and an inlet flow meter (13) are installed at the inlet of the pipeline to be tested, and are used to detect the pressure value and flow rate of the medium at the inlet of the pipeline to be tested (8); an outlet pressure sensor (15) and an outlet flow meter (16) are installed at the outlet of the pipeline to be tested (8), and are used to detect the pressure value and flow rate of the medium at the outlet of the pipeline to be tested; a node pressure sensor is installed at a preset node of the pipeline to be tested, and is used to detect the pressure value of the medium at the preset node of the pipeline to be tested; The output ends of the inlet pressure sensor (12), the inlet flow meter (13), the node pressure sensor, the outlet pressure sensor (15) and the outlet flow meter (16) are all connected to the input end of the acquisition card (17), and the output end of the acquisition card (17) is connected to the input end of the particle filter (18); Wherein, the number of node pressure sensors is one or more; It also comprises: a first auxiliary pipeline (7) and a second auxiliary pipeline (9); the inlet of the pipeline to be tested (8) is connected to the outbound oil storage tank (1) through the first auxiliary pipeline (7), and the outlet of the pipeline to be tested (8) is connected to the inbound oil tank (11) through the second auxiliary pipeline (9); the first ball valve (2), a pump (3), a check valve (4), a second ball valve (5) and a control valve (6) are sequentially arranged on the first auxiliary pipeline (7) along the medium conveying direction; the second auxiliary pipeline (9) is provided with a third ball valve (10); the first auxiliary pipeline (7) and the second auxiliary pipeline (9) are of the same specifications as the pipeline to be tested (8); The particle filter (18) comprises: a wireless transmission module, a data storage module, a calculation module and a display module; The receiving end of the wireless transmission module is used to receive data transmitted by the acquisition card (17); the output end of the wireless transmission module is connected to the receiving end of the data storage module; the output end of the data storage module is connected to the input end of the calculation module; the output end of the calculation module is connected to the input end of the display module; the calculation module is a single chip microcomputer with a preset calculation capacity, and a preset particle filter algorithm program is integrated in the single chip microcomputer; The positioning method comprises the following steps: S1, collecting and acquiring basic data and working data of the pipeline to be tested; the working data at least includes the inlet pressure value and flow value of the pipeline to be tested, the outlet pressure value and flow value, and the pressure value at a preset node in the pipeline to be tested; S2, according to the data collected in step S1, using a transient model method based on adding leakage information to establish a state space equation of the pipeline fluid; S3, judging whether leakage occurs by the flow value collected in step S1; if leakage occurs, outputting the preliminary leakage amount and jumping to step S4; if leakage does not occur, jumping to step S1; S4, according to the state space equation obtained in step S2, the particle filter and the preliminary leakage amount are initialized, and the pipeline state at the time of leakage is preliminarily estimated based on the particle filter algorithm, that is, all preset nodes are traversed in a preset order, and a set of estimated values ​​whose difference between the estimated value and the observed value is less than a preset threshold value is found, and the leakage amount, leakage coefficient and preliminary leakage position of the pipeline to be tested are obtained; S5, by increasing the number of preset nodes, a state space equation that is more accurate than step S2 is established, and then the particle filter is initialized. The leakage position and leakage amount in the initialization condition are obtained by step S4, and the pipeline state at the time of leakage is estimated based on the particle filter algorithm. When the difference between the estimated value and the observed value is less than a preset threshold that is more accurate than step S4, the optimized leakage position is obtained.

2. A pipeline leakage detection and leakage location positioning method according to claim 1, characterized in that: Step S1 specifically includes: collecting the length L of the pipeline to be measured, the inner diameter D of the pipeline to be measured, the hydraulic friction coefficient f, the pressure wave propagation speed C s , the horizontal angle of the pipeline θ; collect the pressure head value H1 of the pipeline inlet oil transfer station, the flow value Q1 and the pressure head value H1 of the outlet oil receiving station n , flow value Q n ; Collect the pressure head value H at the preset node position of the pipeline i′ ; Collect the density ρ of the transport medium in the pipeline.

3. A pipeline leakage detection and leakage location positioning method according to claim 2, characterized in that: Step S2 specifically includes: establishing a mathematical model of the pipeline fluid through mass balance and momentum balance of the transported medium, and using the characteristic line method to obtain a numerical solution; The expression of the pipeline state space equation including leakage is constructed as follows: x j+1 =Ax j +Bu j +Not j 2 +Eu j+1 +aξ+FQ l 4 y=Cx j In the formula, the state of the pipe flow system is x = [H1, ..., H n-1 , Q2, …, Q n ] T , j represents the time step of the pipeline system, n is the number of nodes in the pipeline, u j is the control quantity of the system, λ l is the leakage coefficient of the leakage point; H l is the pressure head value at the leakage point, y is the measured value, and the output matrix C is: Q l The actual leakage amount of the flow meter.

4. A pipeline leakage detection and leakage location positioning method according to claim 3, characterized in that: Step S3 specifically includes: Set the threshold th1 and the response delay t; when |Q1-Q n |>th1, and this state lasts for Δt>t, then it is judged that the pipeline is leaking; otherwise, it is judged that the pipeline is not leaking; when a leak occurs, the leakage time t′ is output, and the average leakage amount within Δt is output. Among them, Q1 is the flow value detected at the pipeline inlet, Q n The flow rate value detected at the pipe outlet.

5. A pipeline leakage detection and leakage location positioning method according to claim 4, characterized in that: Step S4 specifically includes: S4.1: Initialize the parameters in the particle filter algorithm: Initialize the matrix F, add the mean leakage value along the pipeline direction in the n-2 node positions where the pipeline may leak Initialize the state space equation, set time j = 0, leakage time j = T'; input the working state of the pipeline transport medium as the control quantity u of the state space equation j ,u j =[Q 1,j ,H n,j ] T ; Set the number of particles to P, initialize the pipeline state particle set χ and the leakage particle set τ; where the mean is Generate the leakage particle set τ using the normal distribution of; S4.2: The initial state x of the pipeline from step S4.1 j , control quantity u j ,u j+1 Calculate the pipeline status x j+1 Based on the particle set x in step S4.1 j , τ j And the observed value y=[H 1,j ,Q n,j ] T , calculate and obtain the particle set χ j+1 ; Calculate x at time j j+1 With χ j+1 The difference dis(i) is calculated as follows: dis(i)=x j+1 -x j+1 Calculate the particle set χ j+1 The weights w(i) corresponding to the P particles in the equation are normalized to q(i). The calculation formula is: Where i is the calculation step required for importance resampling, i∈[1,P]; j+1 and weight value q(i), we get a set of weighted particle sets The posterior density function that approximately represents the pipeline state at time j+1, where the weights satisfy the normalization condition S4.3: For the weighted particle set χ′ obtained in step S4.2 j+1 Using the importance resampling method, χ′ j+1 After the particles with lower weights are replaced by particles with higher weights, they are used as the particle set to estimate the state of the pipeline at time j+2; before the particle filter estimation at time j+2, the particle set χ′ j+1 Update the updated particle set τ j Perform pre-processing; S4.4: Repeat steps S4.2 and S4.3 to obtain the particle filter estimation value of the pipeline system at any time of leakage state Particle filter estimate of leakage S4.5: Compare measured values ​​with estimated values ​​at each node in the piping system Compare the measured leakage Q l With estimated value In step S4.1, determine whether the leakage information attached to the node n along the pipeline is correct; if it is correct, the attachment of the leakage information is valid, and the estimated leakage position is output. If it is incorrect, update the leakage position n=n+1 in step S4.1 and repeat steps S4.2 to S4.

5.

6. A pipeline leakage detection and leakage location positioning method according to claim 5, characterized in that: Step S5 specifically includes: Step S5.1: increasing the number of segments N′ of the pipeline to be tested in step S1 to establish a more accurate state space equation of the pipeline to be tested; Step S5.2: Set a threshold value th2 that is more accurate than that in step S4. The additional leakage position and leakage amount in the initialization condition of the particle filter are the values ​​obtained in step S4. The leakage position The update method is Where Δx′ is the distance between two nodes of the state space equation in step S5.1, and the value range of m is To round up to an integer; Iterate and update until the preset threshold condition is met, and output accurate leakage location information Leakage coefficient λ of the leak hole l .

7. A pipeline leakage detection and leakage location positioning method according to claim 5, characterized in that: Step S4.3 includes: (1) Generate a constant in the interval [0, 1] based on the normal distribution, denoted as r(i); (2) The weight of the pipeline state at this moment obtained from step 4.2 Compare with r(i); if Update k = k + 1 to get Continue to compare with r(i); if The particle with the higher weight corresponding to weight q(i=k) Replace particles with lower weights (3) According to the number of particles P in the particle filter, repeat steps (1) and (2) for P times; and Take the mean value to obtain the particle filter estimate of the pipeline state at time j+1 Particle filter estimation of leakage at time j+1 (4) Update particle set χ j+1 = x" j+1 , update the particle set τ j+1 =τ′ j+1 ; The specific preprocessing method is: τ′ j+1 (i) = τ′ j+1 (i)+δ(i), where δ(i) obeys a normal distribution with a mean of 0, and the required variance of the normal distribution depends on the actual leakage amount.

Images