Leakage anomaly point accurate positioning method, system and product based on artificial electric field method

Abstract

This invention relates to the field of leakage detection technology, providing a method, system, and product for precise location of leakage anomalies based on artificial electric field method. The method includes: acquiring the coordinates and potential values ​​of sensors arranged in an artificial electric field; calculating the minimum distance between each sensor; generating potential distribution data of the artificial electric field based on Kriging interpolation; detecting multiple peaks using the minimum distance between sensors as the detection range; selecting N closest sensors for each peak, and then selecting M sensors from these M sensors, ensuring that the M sensors are non-collinear and have the largest sum of potential values; calculating the potential objective function for each peak based on the M sensors; locating the initial leakage point using various optimization methods based on the potential objective function for each peak; and detecting anomalies at the initial leakage point and calculating the average value to obtain the final leakage anomaly point. This invention optimizes the sensor layout, improving leakage location accuracy and the robustness of the monitoring system.

Description

Technical Field

[0001] This invention relates to the field of leakage detection technology, and in particular to a method, system and product for accurately locating leakage anomalies based on artificial electric field method. Background Technology

[0002] Artificial electric field method is a widely used non-destructive testing technique in the field of leakage detection. It involves applying an artificial electric field to the detection area, causing the leakage channel to form an equivalent conductive anomaly. Sensors are then deployed to collect spatial potential distribution data, and the leakage location is determined based on the potential gradient and anomaly characteristics. However, traditional leakage anomaly point location relies on contour maps, which cover large areas and cannot accurately pinpoint the leakage location, leading to difficult and costly repairs. Traditional sensors are often numerous and spaced closely together, making construction difficult and resulting in a high failure rate. Conversely, fewer sensors lead to low leakage location accuracy and difficulty. The lack of a systematic approach to sensor selection further hinders optimization, resulting in decreased leakage location accuracy. Traditional techniques typically detect single, large leakage anomalies but struggle to detect smaller, surrounding anomalies, making it difficult to handle multiple anomalies. Furthermore, traditional techniques lack adequate outlier handling, and the detection data is easily distorted by occasional signal interference, affecting the final location accuracy. Existing artificial electric field leakage detection technology has significant shortcomings in terms of positioning accuracy, sensor optimization, multi-anomaly identification, anti-interference, and robustness. It cannot simultaneously achieve the comprehensive goals of "low-cost deployment, high-precision positioning, multi-leakage identification, and highly robust operation," thus hindering its widespread application in high-precision engineering leakage detection. Therefore, there is an urgent need for a leakage location method that can optimize sensor layout, accurately locate leakage anomalies, adapt to multiple leakage scenarios, and possess strong anti-interference capabilities. Summary of the Invention

[0003] This invention aims to at least solve one of the technical problems existing in related technologies. To this end, this invention provides a method, system, and product for precise location of leakage anomalies based on the artificial electric field method, optimizing sensor layout, improving leakage location accuracy, and enhancing the robustness of the monitoring system.

[0004] This invention provides a method for accurately locating leakage anomalies based on an artificial electric field, comprising:

[0005] S1: Collect the sensor coordinates and sensor potential values ​​of the artificial electric field arrangement of the liquid storage equipment, and calculate the minimum distance between each sensor based on the sensor coordinates;

[0006] S2: Based on the coordinates and potential values ​​of each sensor, artificial electric field potential distribution data is generated using the Kriging interpolation method. The minimum distance between each sensor is used as the search radius to detect multiple peaks in the artificial electric field potential distribution data.

[0007] S3: For each peak, select the N closest sensors, and select M sensors from the N sensors. The M sensors are not collinear and have the largest sum of potential values. Calculate the potential objective function for each peak based on the potential values ​​of the M sensors.

[0008] S4: Based on the potential objective function of each peak, the initial leakage point is located by a combination optimization strategy of L-BFGS-B algorithm, Nelder-Mead algorithm, Powell algorithm and simplified location method;

[0009] S5: Perform outlier detection on the initial leakage point to obtain the filter leakage point, calculate the average value of all filter leakage points, and obtain the final leakage anomaly point of the liquid storage equipment.

[0010] Furthermore, the distance between the two sensors is calculated using Euclidean distance.

[0011] Furthermore, the artificial electric field potential distribution data generated based on the Kriging interpolation method includes:

[0012] S211: Determine the detection area based on the sensor coordinates arranged according to the artificial electric field of the liquid storage device, and generate grid points around the sensor based on the maximum boundary of the detection area;

[0013] S212: Calculate the variogram values ​​of grid points around the sensor using a spherical variogram model;

[0014] S213: Construct a set of Kriging equations based on the variogram values ​​of the grid points around the sensors, solve the set of Kriging equations, and obtain the weights of each sensor;

[0015] S214: Calculate the potential value of each grid point based on the weight of each sensor and the potential value of each sensor, and generate the potential distribution around the sensor based on the potential value of each grid point.

[0016] Furthermore, using the minimum distance between each sensor as the search radius, multiple peaks in the artificial electric field potential distribution data are detected, including:

[0017] S221: Traverse each sensor and check the potential value of each grid point within a range centered on the sensor and with the minimum distance between the sensors as the radius;

[0018] S222: For each sensor, compare the sensor's potential value with the maximum potential value of all grid points within the minimum distance range between the surrounding sensors; if the sensor's potential value is greater than the maximum potential value of all grid points within the minimum distance range, then the sensor is a leakage abnormal peak point and is added to the peak candidate list.

[0019] S223: Verify the peak candidate list. If the peak candidate list contains the point with the largest potential value in the detection area, and all peak values ​​in the peak candidate list are greater than or equal to the preset peak threshold, then the peak candidate list is the final peak list. If the peak candidate list does not contain the point with the largest potential value in the detection area, then the point with the largest potential value in the detection area is added to the peak candidate list to generate the final peak list.

[0020] Furthermore, step S3 includes:

[0021] S31: For each peak value, calculate the distance from all sensors to the peak value, and select the N sensors that are closest to the peak value;

[0022] S32: Sort the N nearest sensors in descending order of their potential values, and select the M non-collinear sensors with the largest sum of their potential values.

[0023] Furthermore, step S4 includes:

[0024] S41: Set the sensor enclosure range as the boundary constraint of the L-BFGS-B algorithm;

[0025] S42: Input the sensor coordinates and sensor potential values ​​into the L-BFGS-B algorithm with boundary constraints of the sensor-enclosed area. If the potential objective function of each peak meets the preset threshold, the L-BFGS-B algorithm outputs the initial leakage point and executes step S5.

[0026] If the potential objective function of each peak does not meet the preset threshold, a penalty term is added to the potential objective function of each peak; proceed to step S43.

[0027] S43: Input the sensor coordinates and sensor potential values ​​into the Nelder-Mead algorithm. If the potential objective function with added penalty terms for each peak meets the preset threshold, the Nelder-Mead algorithm outputs the initial leakage point and executes step S5.

[0028] If the potential objective function with added penalty terms for each peak does not meet the preset threshold, then the sensor coordinates and sensor potential values ​​are input into the Powell algorithm.

[0029] If the potential objective function with added penalty terms for each peak meets the preset threshold, the Powell algorithm outputs the initial leakage point and executes step S5.

[0030] If the potential objective function with added penalty terms for each peak does not meet the preset threshold, then the sensor coordinates and sensor potential values ​​are input into the simplified positioning method, the initial leak point is output, and step S5 is executed.

[0031] Furthermore, the simplified positioning method uses the sensor location with the largest potential value as the initial leak point.

[0032] Furthermore, in step S5, outlier detection is performed on the initial leak point using the IQR method.

[0033] This invention also provides a system for accurately locating leakage anomalies based on an artificial electric field method, used to execute the aforementioned method for accurately locating leakage anomalies based on an artificial electric field method, comprising:

[0034] The data acquisition module acquires the sensor coordinates and sensor potential values ​​of the artificial electric field arrangement of the liquid storage device, and calculates the minimum distance between each sensor based on the sensor coordinates.

[0035] An interpolation module generates artificial electric field potential distribution data based on the coordinates and potential values ​​of each sensor and the Kriging interpolation method. The module uses the minimum distance between each sensor as the search radius to detect multiple peaks in the artificial electric field potential distribution data.

[0036] A peak selection module selects N sensors that are closest to each peak, selects M sensors from the N sensors, the M sensors are not collinear and have the largest sum of potential values, and calculates the potential objective function for each peak based on the potential values ​​of the M sensors.

[0037] The positioning module locates the initial leakage point based on the potential objective function of each peak using a combination optimization strategy of the L-BFGS-B algorithm, Nelder-Mead algorithm, Powell algorithm, and simplified positioning method.

[0038] The filtration and calculation module detects outliers at the initial leakage points, obtains the filtration leakage points, calculates the average value of all filtration leakage points, and obtains the final leakage anomaly points of the liquid storage equipment.

[0039] The present invention also provides a computer program product, including a computer program that, when executed by a processor, implements the above-described method for accurately locating leakage anomalies based on an artificial electric field.

[0040] The above-described one or more technical solutions in the embodiments of the present invention have at least one of the following technical effects:

[0041] This invention, based on traditional contour mapping, further determines the location of leakage anomalies, narrows down the suspected area of ​​leakage anomalies, and improves the accuracy of leakage anomaly location. The invention designs sensor spacing according to the actual application site and uses a multi-peak judgment algorithm to avoid missed detections and location deviations. It calculates leakage anomaly location using data from multiple datasets and time spans, eliminating anomalies with significant deviations and removing interference from random signals. Multiple optimization methods and fault-tolerant mechanisms are employed to ensure a high success rate of location in various scenarios. It reduces manual intervention and is suitable for real-time monitoring. It can be applied to leakage detection in various scenarios such as water tanks, reservoirs, dams, etc.

[0042] Additional aspects and advantages of the invention will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of the invention. Attached Figure Description

[0043] To more clearly illustrate the technical solutions in this invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0044] Figure 1 This is a flowchart illustrating a method for accurately locating leakage anomalies based on an artificial electric field, provided by the present invention.

[0045] Figure 2 This is a schematic diagram of a layout with four sensors arranged only at the corners, provided in an embodiment of the present invention.

[0046] Figure 3 This embodiment of the invention uses the traditional contour line detection method to detect the detection cloud map of four sensors.

[0047] Figure 4 This is a schematic diagram illustrating the positioning results of detecting four sensors using this invention in an embodiment of the invention.

[0048] Figure 5 This is a schematic diagram of the layout of 18 sensors arranged within the detection range provided in an embodiment of the present invention.

[0049] Figure 6 This is a detection cloud map of 18 sensors detected by the conventional contour line detection method in an embodiment of the present invention.

[0050] Figure 7 This is a schematic diagram illustrating the positioning results of 18 sensors detected using this invention in an embodiment of the invention.

[0051] Figure 8This is a schematic diagram of the structure of a system for accurately locating leakage anomalies based on the artificial electric field method provided by the present invention.

[0052] Figure label:

[0053] 101. Acquisition Module; 102. Interpolation Module; 103. Peak Selection Module; 104. Positioning Module; 105. Filtering and Calculation Module. Detailed Implementation

[0054] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of this invention will be clearly and completely described below. Obviously, the described embodiments are only some, not all, of the embodiments of this invention. Based on the embodiments of this invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this invention. The following embodiments are used to illustrate this invention but cannot be used to limit the scope of this invention.

[0055] In the description of this specification, the references to terms such as "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., refer to specific features, structures, materials, or characteristics described in connection with that embodiment or example, which are included in at least one embodiment or example of the present invention. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples. Moreover, without contradiction, those skilled in the art can combine and integrate the different embodiments or examples described in this specification, as well as the features of different embodiments or examples.

[0056] The following is combined Figures 1 to 8 This invention describes a method for accurately locating leakage anomalies based on an artificial electric field, comprising:

[0057] S1: Collect the sensor coordinates and sensor potential values ​​of the artificial electric field arrangement of the liquid storage equipment, and calculate the minimum distance between each sensor based on the sensor coordinates;

[0058] Collect the sensor coordinates and sensor potential values ​​of the artificial electric field arrangement of the liquid storage equipment, and verify the integrity of the data to ensure that the sensor coordinates and sensor potential values ​​are valid;

[0059] Multiple sensors are arranged in a grid pattern around the detection area; an artificial electric field is applied to make the leak point form an equivalent point charge; the sensors measure the potential value at their location; the data is stored as a CSV file, which includes the sensor number, sensor X coordinate, sensor Y coordinate, and the potential value measured by the sensor.

[0060] The verification data is free of missing sensor numbers, sensor X coordinates, sensor Y coordinates, and sensor measured potential values ​​to ensure that the data is within the reasonable range for engineering applications and to exclude outliers and erroneous values.

[0061] The distance between the two sensors is calculated using Euclidean distance, and the expression is as follows:

[0062]

[0063] in, For the first The sensor and the first Distance between sensors For the first Sensor coordinates, For the first Sensor coordinates.

[0064] By comparing the distances between each pair of sensors and selecting the minimum value, the minimum distance between each sensor is obtained.

[0065] S2: Based on the coordinates and potential values ​​of each sensor, artificial electric field potential distribution data is generated using the Kriging interpolation method. The minimum distance between each sensor is used as the search radius to detect multiple peaks in the artificial electric field potential distribution data.

[0066] The artificial electric field potential distribution data generated based on the Kriging interpolation method includes:

[0067] S211: Determine the detection area based on the sensor coordinates arranged according to the artificial electric field of the liquid storage device, and generate grid points around the sensor based on the maximum boundary of the detection area;

[0068] Based on the coordinates of all sensors, the maximum boundary of the detection area is determined, and a preset buffer distance is extended outward, such as 1 / 2 of the minimum spacing between sensors, to form a complete calculation boundary. The area is uniformly divided using a Cartesian coordinate system, the grid step size is set, a structured grid point set covering the entire detection area is generated, and the planar coordinates of each grid point are recorded.

[0069] S212: Calculate the variogram values ​​of grid points around the sensor using a spherical variogram model;

[0070] Generate the potential distribution around the sensor; the calculation expression is:

[0071]

[0072] in, The value of the variogram. Value of a nugget. For sill values, For spatial distance, For variable range.

[0073] S213: Construct a set of Kriging equations based on the variogram values ​​of the grid points around the sensors, solve the set of Kriging equations, and obtain the weights of each sensor;

[0074] S214: Calculate the potential value of each grid point based on the weight of each sensor and the potential value of each sensor, and generate the potential distribution around the sensor based on the potential value of each grid point;

[0075] The formula for calculating the potential value at each grid point is:

[0076]

[0077] in, Let the potential at the grid points be... For the first Sensor weights, For the first The potential of each sensor, This represents the number of sensors.

[0078] Using the minimum distance between each sensor as the search radius, multiple peaks in the artificial electric field potential distribution data are detected, including:

[0079] S221: Traverse each sensor and check the potential value of each grid point within a range centered on the sensor and with the minimum distance between the sensors as the radius;

[0080] Starting from the location of all sensors, check the coordinates and corresponding potential values ​​of all grid points within the detection range of each sensor, with the minimum distance between each sensor as the radius.

[0081] S222: For each sensor, compare the sensor's potential value with the maximum potential value of all grid points within the minimum distance range between the surrounding sensors; if the sensor's potential value is greater than the maximum potential value of all grid points within the minimum distance range, then the sensor is a leakage abnormal peak point and is added to the peak candidate list;

[0082] S223: Verify the peak candidate list. If the peak candidate list contains the point with the largest potential value in the detection area, and all peak values ​​in the peak candidate list are greater than or equal to a preset peak threshold (when a peak value is less than the preset threshold, it will be removed), then the peak candidate list is the final peak list. If the peak candidate list does not contain the point with the largest potential value in the detection area, then the point with the largest potential value in the detection area is added to the peak candidate list to generate the final peak list.

[0083] For sensors If its potential value satisfy:

[0084]

[0085] in, For sensors All grid points within a range of d. To find the function with the maximum value, For grid points Electric potential value.

[0086] S3: For each peak, select the N closest sensors, and select M sensors from the N sensors. The M sensors are not collinear and have the largest sum of potential values. Calculate the potential objective function for each peak based on the potential values ​​of the M sensors.

[0087] S31: For each peak value, calculate the distance from all sensors to the peak value, and select the N sensors that are closest to the peak value;

[0088] S32: Sort the N nearest sensors in descending order of their potential values, and select the M non-collinear sensors with the largest sum of their potential values.

[0089] In some specific embodiments of the present invention, N=5 and M=4.

[0090] The expression for calculating the objective function of the potential for each peak is:

[0091]

[0092] in, Let the electric potential be the objective function. The actual potential of the sensor. Let C be the distance from the sensor to the point to be determined, and C be the point charge intensity. The objective function is to find the minimum value.

[0093] S4: Based on the potential objective function of each peak, the initial leakage point is located by a combination optimization strategy of L-BFGS-B algorithm, Nelder-Mead algorithm, Powell algorithm and simplified location method;

[0094] S41: Set the sensor enclosure range as the boundary constraint of the L-BFGS-B algorithm;

[0095] S42: Input the sensor coordinates and sensor potential values ​​into the L-BFGS-B algorithm with boundary constraints of the sensor-enclosed area. If the potential objective function of each peak meets the preset threshold, the L-BFGS-B algorithm outputs the initial leakage point and executes step S5.

[0096] If the potential objective function of each peak does not meet the preset threshold, a penalty term is added to the potential objective function of each peak, and step S43 is executed.

[0097] L-BFGS-B is a quasi-Newton method with boundary constraints. It is an improved version of the BFGS (Broyden-Fletcher-Goldfarb-Shanno) algorithm. It approximates the Hessian matrix (second derivative matrix) by storing limited historical information, thereby reducing memory consumption and making it suitable for large-scale optimization problems.

[0098] The optimized calculation for leakage location in this invention ensures that the initial leakage point is within a reasonable range by setting the sensor enclosure range as a boundary constraint.

[0099] S43: Input the sensor coordinates and sensor potential values ​​into the Nelder-Mead algorithm. If the potential objective function with added penalty terms for each peak meets the preset threshold, the Nelder-Mead algorithm outputs the initial leakage point; execute step S5.

[0100] If the potential objective function with added penalty terms for each peak does not meet the preset threshold, then the sensor coordinates and sensor potential values ​​are input into the Powell algorithm.

[0101] If the potential objective function with added penalty terms for each peak meets the preset threshold, the Powell algorithm outputs the initial leakage point and executes step S5.

[0102] If the potential objective function with added penalty terms for each peak does not meet the preset threshold, then the sensor coordinates and sensor potential values ​​are input into the simplified positioning method, the initial leak point is output, and step S5 is executed.

[0103] Nelder-Mead is an unconstrained optimization algorithm based on the simplex method. It works by constructing a simplex matrix... A simplex composed of points (such as a triangle in two-dimensional space) can be adjusted in shape through operations such as reflection, expansion, and contraction to gradually approach the optimal solution.

[0104] Used when L-BFGS-B fails, it handles boundary constraints by adding a penalty term to the objective function.

[0105] Powell's algorithm is an unconstrained optimization algorithm, belonging to the conjugate direction method. It constructs a set of conjugate directions and performs a one-dimensional search along these directions in each iteration to gradually approach the optimal solution.

[0106] Used when both the L-BFGS-B and Nelder-Mead algorithms fail, this algorithm handles boundary constraints through a penalty term in the objective function.

[0107] A simplified location method uses the sensor location with the highest potential value as the initial leak point.

[0108] The L-BFGS-B algorithm is used for boundary constraint optimization. If it fails, the Nelder-Mead algorithm is used for gradient-free optimization. If it fails again, the Powell algorithm is used for conjugate direction optimization. If all of these fail, the sensor position with the maximum potential value is used as the initial positioning result.

[0109] L-BFGS-B (with boundary constraints) is used to take advantage of its fast convergence speed.

[0110] If L-BFGS-B fails, Nelder-Mead is used, taking advantage of its robustness.

[0111] If Nelder-Mead also fails, use Powell as a last resort.

[0112] If all methods fail, use the simplified method.

[0113] This combined strategy significantly improves the robustness of the localization algorithm, ensuring successful location of leakage anomalies in various scenarios.

[0114] S5: Perform outlier detection on the initial leakage point to obtain the filter leakage point, calculate the average value of all filter leakage points, and obtain the final leakage point of the liquid storage equipment.

[0115] The IQR (Interquartile Range) method was used to detect and remove abnormal location results.

[0116] For each peak, the interquartile range is calculated from multiple positioning results. The calculation expression is as follows:

[0117]

[0118] in, This is the interquartile range value. It is the 25th percentile. It is the 75th percentile;

[0119] The outlier threshold is set, and the calculation expression is as follows:

[0120]

[0121] in, The threshold for outliers;

[0122] Remove initial leak points that exceed the outlier threshold to obtain filtered leak points.

[0123] For outliers with high statistical concentration, the final location of the leakage anomaly is determined by calculating the average value.

[0124] Calculate the average value of the filtered positioning results, and use the average value as the final location of the leakage anomaly point in the liquid storage equipment. The calculation expression is as follows:

[0125]

[0126] in, To filter out the location of the leak, This indicates the final leak point in the liquid storage equipment. This represents the number of filter leak points.

[0127] like Figure 8 As shown, a system for accurately locating leakage anomalies based on an artificial electric field method is used to execute the aforementioned method for accurately locating leakage anomalies based on an artificial electric field method, comprising:

[0128] The data acquisition module 101 acquires the sensor coordinates and sensor potential values ​​of the artificial electric field arrangement of the liquid storage equipment, and calculates the minimum distance between each sensor based on the sensor coordinates.

[0129] The interpolation module 102 generates artificial electric field potential distribution data based on the coordinates and potential values ​​of each sensor and the Kriging interpolation method. It uses the minimum distance between each sensor as the search radius to detect multiple peaks in the artificial electric field potential distribution data.

[0130] The peak selection module 103 selects the N closest sensors for each peak, selects M sensors from the N sensors, the M sensors are not collinear and have the largest sum of potential values, and calculates the potential objective function for each peak based on the potential values ​​of the M sensors;

[0131] The positioning module 104 locates the initial leakage point based on the potential objective function of each peak using a combination optimization strategy of the L-BFGS-B algorithm, Nelder-Mead algorithm, Powell algorithm, and simplified positioning method.

[0132] The filtration and calculation module 105 performs outlier detection on the initial leakage point, obtains the filtration leakage point, calculates the average value of all filtration leakage points, and obtains the final leakage anomaly point of the liquid storage device.

[0133] Through the collaborative work of the aforementioned modules, the location of leakage anomalies will be further determined based on traditional contour maps, narrowing down the suspected area of ​​leakage anomalies and improving the accuracy of leakage anomaly location. This invention designs sensor spacing according to the actual application site and uses a multi-peak judgment algorithm to avoid missed detections and location deviations. It calculates leakage anomaly location using data from multiple datasets and time spans, eliminating anomalies with significant deviations and removing interference from random signals. Multiple optimization methods and fault-tolerant mechanisms are employed to ensure a high success rate of location in various scenarios. It reduces manual intervention and is suitable for real-time monitoring. It can be applied to leakage detection in various scenarios such as water tanks, reservoirs, and dams.

[0134] On the other hand, the present invention also provides a computer program product, the computer program product including a computer program stored on a non-transitory computer-readable storage medium, the computer program including program instructions, and when the program instructions are executed by a computer, the computer is able to execute the above-mentioned method for accurately locating leakage anomalies based on artificial electric field method.

[0135] To verify the effectiveness of this invention, leakage anomaly points were located using both the present invention and the traditional contour line detection method in two scenarios: one with four sensors arranged only at the corners within the sensor detection range, and the other with 18 sensors arranged throughout the entire sensor detection range. Figure 2 As shown, only four sensors are arranged at the corners within the sensor detection range. These sensors are sparsely distributed with large spacing, simulating an engineering inspection scenario with a small number of sensors and low deployment costs. Traditional detection methods struggle to locate leaks under this layout, highlighting the technical advantage of this invention in accurately locating leaks even with sparse sensor deployment. Figure 2 The sensor shown is detected using the traditional contour line detection method, as shown in the detection cloud map. Figure 3 As shown, the potential contour lines are distributed gently and the gradient characteristics are not obvious, making it impossible to determine the location of leakage anomalies. The actual leakage points are blurred and covered. Traditional contour line detection methods cannot locate leakage anomalies due to the limited number of sensors. Figure 3 The actual location of the leakage anomaly is (0.8, 0.75), and traditional contour lines are basically unable to determine the location of the leakage anomaly.

[0136] according to Figure 2 The sensor shown in the invention locates the abnormal leakage point detection cloud map as follows: Figure 4 As shown, this invention, through Kriging interpolation, multi-peak detection, sensor optimization, multi-level optimization, and outlier removal, finally outputs accurate coordinates of the leakage anomaly point (1.091, 0.778). Compared with the actual leakage point (0.8, 0.75), the positioning error is only 0.292m, which solves the problem that traditional methods have too few sensors to detect leakage anomalies and ensures detection accuracy.

[0137] like Figure 5 The image shows 18 sensors arranged in a grid pattern within the sensor detection range. The large number of sensors with small spacing are used to simulate a high-density detection scenario, and to compare the positioning accuracy difference between traditional methods and the present invention under dense sensor conditions.

[0138] according to Figure 5 The sensor shown is detected using the traditional contour line detection method, as shown in the detection cloud map. Figure 6 As shown, the leakage area was roughly determined by contour lines, and two abnormal leakage points were obtained: the first one was (0,0) and the second one was (6.4,0). The actual leakage points were (0.8,0.35) and (7.2,0.35) respectively, with an error of 0.873m for both. The positioning deviation was large and the accuracy was low, which could not meet the requirements of precise repair in engineering.

[0139] according to Figure 5 The sensor shown in the invention locates the abnormal leakage point detection cloud map as follows: Figure 7 As shown, this invention, through Kriging interpolation, multi-peak detection, sensor optimization, multi-level optimization, and outlier removal, ultimately obtains two leakage anomaly points: the first at (0.610, 0.189) and the second at (6.871, 0.269). The actual leakage point locations are (0.8, 0.35) for the first and (7.2, 0.35) for the second, with errors of 0.249m and 0.339m respectively. These represent improvements in accuracy of approximately 71.5% and 61.2% compared to traditional methods. This invention can simultaneously identify multiple leakage points, significantly improving positioning accuracy and outperforming traditional detection methods.

[0140] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for accurately locating leakage abnormal points based on the artificial electric field method, characterized in that, include: S1: Collect the sensor coordinates and sensor potential values ​​of the artificial electric field arrangement of the liquid storage equipment, and calculate the minimum distance between each sensor based on the sensor coordinates; S2: Based on the coordinates and potential values ​​of each sensor, artificial electric field potential distribution data is generated using the Kriging interpolation method. The minimum distance between each sensor is used as the search radius to detect multiple peaks in the artificial electric field potential distribution data. S3: For each peak, select the N closest sensors, and select M sensors from the N sensors. The M sensors are not collinear and have the largest sum of potential values. Calculate the potential objective function for each peak based on the potential values ​​of the M sensors. S4: Based on the potential objective function of each peak, the initial leakage point is located by a combination optimization strategy of L-BFGS-B algorithm, Nelder-Mead algorithm, Powell algorithm and simplified location method; S5: Perform outlier detection on the initial leakage point to obtain the filter leakage point, calculate the average value of all filter leakage points, and obtain the final leakage anomaly point of the liquid storage equipment.

2. The method according to claim 1, wherein, The distance between the two sensors is calculated using Euclidean distance.

3. The method according to claim 1, wherein, The artificial electric field potential distribution data generated based on the Kriging interpolation method includes: S211: Determine the detection area based on the sensor coordinates arranged according to the artificial electric field of the liquid storage device, and generate grid points around the sensor based on the maximum boundary of the detection area; S212: Calculate the variogram values ​​of grid points around the sensor using a spherical variogram model; S213: Construct a set of Kriging equations based on the variogram values ​​of the grid points around the sensors, solve the set of Kriging equations, and obtain the weights of each sensor; S214: Calculate the potential value of each grid point based on the weight of each sensor and the potential value of each sensor, and generate the potential distribution around the sensor based on the potential value of each grid point.

4. The method according to claim 1, wherein, Using the minimum distance between each sensor as the search radius, multiple peaks in the artificial electric field potential distribution data are detected, including: S221: Traverse each sensor and check the potential value of each grid point within a range centered on the sensor and with the minimum distance between the sensors as the radius; S222: For each sensor, compare the sensor's potential value with the maximum potential value of all grid points within the minimum distance range between the surrounding sensors; if the sensor's potential value is greater than the maximum potential value of all grid points within the minimum distance range, then the sensor is a leakage abnormal peak point and is added to the peak candidate list. S223: Verify the peak candidate list. If the peak candidate list contains the point with the largest potential value in the detection area, and all peak values ​​in the peak candidate list are greater than or equal to the preset peak threshold, then the peak candidate list is the final peak list. If the peak candidate list does not contain the point with the largest potential value in the detection area, then the point with the largest potential value in the detection area is added to the peak candidate list to generate the final peak list.

5. The method for accurately locating leakage anomalies based on artificial electric field method according to claim 1, characterized in that, Step S3 includes: S31: For each peak value, calculate the distance from all sensors to the peak value, and select the N sensors that are closest to the peak value; S32: Sort the N nearest sensors in descending order of their potential values, and select the M non-collinear sensors with the largest sum of their potential values.

6. The method for accurately locating leakage anomalies based on artificial electric field method according to claim 1, characterized in that, Step S4 includes: S41: Set the sensor enclosure range as the boundary constraint of the L-BFGS-B algorithm; S42: Input the sensor coordinates and sensor potential values ​​into the L-BFGS-B algorithm with boundary constraints of the sensor-enclosed area. If the potential objective function of each peak meets the preset threshold, the L-BFGS-B algorithm outputs the initial leakage point and executes step S5. If the potential objective function of each peak does not meet the preset threshold, a penalty term is added to the potential objective function of each peak, and step S43 is executed. S43: Input the sensor coordinates and sensor potential values ​​into the Nelder-Mead algorithm. If the potential objective function with added penalty terms for each peak meets the preset threshold, the Nelder-Mead algorithm outputs the initial leakage point and executes step S5. If the potential objective function with added penalty terms for each peak does not meet the preset threshold, then the sensor coordinates and sensor potential values ​​are input into the Powell algorithm. If the potential objective function with added penalty terms for each peak meets the preset threshold, the Powell algorithm outputs the initial leakage point and executes step S5. If the potential objective function with added penalty terms for each peak does not meet the preset threshold, then input the sensor coordinates and sensor potential values ​​into the simplified positioning method, output the initial leak point, and execute step S5.

7. The method for accurately locating leakage anomalies based on artificial electric field method according to claim 1, characterized in that, The simplified positioning method uses the sensor location with the highest potential value as the initial leak point.

8. The method for accurately locating leakage anomalies based on artificial electric field method according to claim 1, characterized in that, In step S5, outlier detection is performed on the initial leak point using the IQR method.

9. A precise leakage anomaly location system based on the artificial electric field method, characterized in that, To perform the method for accurately locating leakage anomalies based on the artificial electric field method as described in any one of claims 1 to 8, comprising: The data acquisition module acquires the sensor coordinates and sensor potential values ​​of the artificial electric field arrangement of the liquid storage device, and calculates the minimum distance between each sensor based on the sensor coordinates. An interpolation module generates artificial electric field potential distribution data based on the coordinates and potential values ​​of each sensor and the Kriging interpolation method. The module uses the minimum distance between each sensor as the search radius to detect multiple peaks in the artificial electric field potential distribution data. A peak selection module selects N sensors that are closest to each peak, selects M sensors from the N sensors, the M sensors are not collinear and have the largest sum of potential values, and calculates the potential objective function for each peak based on the potential values ​​of the M sensors. The positioning module locates the initial leakage point based on the potential objective function of each peak using a combination optimization strategy of the L-BFGS-B algorithm, Nelder-Mead algorithm, Powell algorithm, and simplified positioning method. The filtration and calculation module detects outliers at the initial leakage points, obtains the filtration leakage points, calculates the average value of all filtration leakage points, and obtains the final leakage anomaly points of the liquid storage equipment.

10. A computer program product, comprising a computer program, characterized in that, When executed by a processor, the computer program implements the method for precise location of leakage anomalies based on artificial electric field method as described in any one of claims 1 to 8.

Images