A method and system for detecting defects in a storage tank roof

Abstract

The application discloses a kind of storage tank roof defect detection method and system. Including: in the storage tank roof arrangement multiple groups of sensors, each group of sensors includes a double element sensor and an excitation sensor, double element sensor includes two receiving sensors, to receive signal from the same group excitation sensor is sent;According to the maximum detection distance of pre-set, select time window length;With Pearson correlation coefficient method respectively determine the similarity of the signal received by the two receiving sensors in each group of sensors in time window;According to the similarity, respectively determine the corresponding damage factor;According to the size of pre-set damage factor influence area and the position relationship of the test point and the corresponding sensor group, determine the weight of the test point about damage factor;According to damage factor and the weight, determine the damage probability of test point. Can expand the effective detection range of single group of sensors, improve detection accuracy, reduce detection cost, reduce the interference of coupling effect between sensors to detection result.

Description

Technical Field

[0001] This invention relates to the field of structural defect detection technology, and in particular to a method and system for detecting defects on the top of a storage tank. Background Technology

[0002] Large storage tanks are the most important facilities in strategic petroleum reserves. These tanks primarily store flammable and explosive liquids. During storage, flammable gases released from the liquids accumulate at the top of the tank. Due to temperature variations, water vapor condenses on the inner wall of the tank top. Oxygen enters the tank through the tank's respiration process and reacts with the liquid on the inner wall of the tank top, causing corrosion. This corrosion leads to structural damage to the tank top, and under the influence of unexpected energy, the flammable gases inside the tank can easily ignite, resulting in an explosion. Due to the large size of the tanks, maintenance costs are high. Generally, conventional non-destructive testing methods for large storage tanks require descaling and rust removal under shutdown conditions, sometimes even requiring opening the tank for inspection. These methods require shutdown, have long inspection cycles, high costs, and high labor intensity. Furthermore, the structural characteristics of the tank top make construction difficult, and the placement of sensors on the top plate is significantly limited, making conventional non-destructive testing impossible. Therefore, when monitoring the health structure of oil storage tank tops, a large-area defect detection method suitable for the top plate is needed to locate corrosion defects.

[0003] Ultrasonic guided waves, with their long propagation distance and sensitivity to small defects, are widely used in the detection of defects in plate-like structures. Existing detection methods primarily rely on comparing the baseline signal from a healthy workpiece with the current signal from a damaged workpiece to obtain the defect signal. While these methods can obtain relatively accurate defect scattering signals in a laboratory environment, in practical applications, factors such as coupling effects, environmental noise, and weather mean that the current signal from a damaged workpiece and the baseline signal from a healthy workpiece are not comparable signals under the same conditions. In recent years, scholars have conducted research in the field of baseline-free defect detection. A. Kumar et al., by time-reversing the received signal and then re-exciting it at the receiving end, propagated the signal back to the excitation end within the plate, and utilized the time reversibility of Lamb waves to distinguish the presence or absence of damage on the metal plate, thus achieving structural health monitoring of isotropic plates. KSAlguri et al. proposed combining the waveguide propagation characteristics and geometric information of the structure under test to construct a comprehensive model similar to the structure under test, forming a dictionary. Damage information is obtained by comparing the current test data with the synthetic baseline data in the dictionary. Yan Shaowen et al. proposed a baseline-free crack identification method based on nonlinear high-frequency response. By utilizing the nonlinear frequency response characteristics under different amplitude excitations, fatigue cracks can be effectively detected. In the above studies, better detection results can be achieved when the defect is distributed on the connection line between the excitation and receiving sensors. To achieve accurate positioning, a large number of dense sensor paths must be arranged to form a sensor network to cover the monitored structure, which increases the detection cost accordingly. Differences in the coupling between different sensors also affect the detection effect. Summary of the Invention

[0004] In order to at least partially solve the above problems, the inventors made this invention, which provides a method and system for detecting defects on the top of storage tanks through specific embodiments.

[0005] In a first aspect, embodiments of the present invention provide a method for detecting defects on the top of a storage tank, comprising the following steps:

[0006] Multiple sets of sensors are arranged on the top of the storage tank. Each set of sensors includes a dual-element sensor and an excitation sensor. The dual-element sensor includes two receiving sensors. The excitation sensor is located on the axis of symmetry of the two receiving sensors. The receiving sensor receives signals emitted from the same set of excitation sensors.

[0007] Select the time window length based on the preset maximum detection distance;

[0008] Using the Pearson correlation coefficient method, the similarity of the signals received by two receiving sensors in each group of sensors within the time window is determined respectively;

[0009] The damage factor of the corresponding sensor group is determined based on the similarity of the signals received by the two receiving sensors in each group.

[0010] Based on the size of the influence area of ​​the damage factor and the positional relationship between the test point and the corresponding sensor group, the weight of the test point with respect to the damage factor is determined.

[0011] The damage probability of the test point is determined based on the damage factor of each sensor group and the weight of the test point with respect to the damage factor.

[0012] Specifically, the maximum detection distance refers to the maximum distance from the test point that a group of sensors can detect to the axis of symmetry of the two receiving sensors in the same group of dual-element sensors.

[0013] Specifically, determining the similarity of signals received by two receiving sensors in each group of sensors within the time window using the Pearson correlation coefficient method includes the following steps:

[0014] The signals received by the two receiving sensors in each group of sensors within the time window are determined and denoted as f. k1 (t) and f k2 (t), where f k1 (t) represents the signal received by one of the receiving sensors in the k-th group of sensors, f k2 (t) represents the signal received by another receiving sensor in the k-th sensor group, where k represents the sensor group number and t represents time;

[0015] The similarity of the signals received by the two receiving sensors in each group of sensors is determined by the following formula:

[0016]

[0017] Where, r k The similarity between the signals received by the two receiving sensors in the k-th group of sensors is represented by T, which represents the length of the time window.

[0018] Specifically, determining the impairment factor for a corresponding sensor group based on the similarity of the signals received by the two receiving sensors in each group includes the following steps:

[0019] The damage factor for each group of sensors is determined using the following formula:

[0020] DI k =1-r k

[0021] Where k is the sensor group number, DI k r represents the damage factor of the k-th sensor group. kThis represents the similarity between the signals received by the two receiving sensors in the k-th group of sensors.

[0022] Specifically, determining the weight of the test point with respect to the damage factor based on the preset size of the area affected by the damage factor and the positional relationship between the test point and the corresponding sensor group includes the following steps:

[0023] The ratio of the sum of the distances from the point to be measured to the midpoints of the excitation sensor and the two receiving sensors in each group of sensors to the distance from the excitation sensor to the midpoints of the two receiving sensors is determined by the following formula:

[0024]

[0025] Where k is the sensor group number, d k Let (x, y) be the distance ratio of the k-th group of sensors, and (x, y) be the coordinates of the point to be measured. ka y ka (x) represents the coordinates of the excitation sensor in the k-th sensor group. ks y ks ) represents the coordinates of the midpoint between the two receiving sensors in the k-th sensor group;

[0026] The weight of the test point with respect to the damage factor is determined by the following formula:

[0027]

[0028] Among them, R k (x,y) represents the weight of the test point (x,y) with respect to the damage factor of the k-th sensor group, γ k γ represents the parameter indicating the region affected by the damage factor of the preset k-th sensor group. k The larger the value, the larger the area affected by the damage factor of the k-th sensor group.

[0029] Specifically, determining the damage probability of a test point based on the damage factor of each sensor group and the weight of the test point with respect to the damage factor includes the following steps:

[0030] The damage probability of the test point is obtained by fusing the damage probabilities of the test point with respect to at least two sets of sensors using the following formula:

[0031]

[0032] Where I(x,y) is the damage probability of the test point (x,y), k is the sensor group number, N is the number of sensor groups participating in the fusion, and DI k R represents the damage factor of the k-th sensor group. k (x,y) represents the weight of the test point (x,y) with respect to the damage factor of the k-th sensor group, DIk R k (x,y) represents the damage probability of the test point (x,y) with respect to the k-th sensor group.

[0033] Optionally, the following steps may also be included:

[0034] The test points whose damage probability exceeds a preset threshold are identified as the locations of the defects.

[0035] Optionally, the following steps may also be included:

[0036] Based on the damage probability of each point on the top of the storage tank, draw a contour map of the damage probability of the top of the storage tank.

[0037] The damage status of the tank top is determined based on the damage probability contour map.

[0038] Secondly, embodiments of the present invention provide a tank top defect detection system, comprising:

[0039] A sensor group arrangement module is used to arrange multiple groups of sensors on the top of a storage tank. Each group of sensors includes a dual-element sensor and an excitation sensor. The dual-element sensor includes two receiving sensors. The excitation sensor is located on the axis of symmetry of the two receiving sensors. The receiving sensor receives signals emitted from the same group of excitation sensors.

[0040] The time window selection module is used to select the length of the time window based on the preset maximum detection distance;

[0041] The similarity determination module is used to determine the similarity of signals received by two receiving sensors in each group of sensors within the time window using the Pearson correlation coefficient method.

[0042] The damage factor determination module is used to determine the damage factor of the corresponding sensor group based on the similarity of the signals received by the two receiving sensors in each group of sensors.

[0043] The damage factor weight determination module is used to determine the weight of the test point with respect to the damage factor based on the size of the preset damage factor influence area and the positional relationship between the test point and the corresponding sensor group.

[0044] The damage probability determination module is used to determine the damage probability of the test point based on the damage factor of each group of sensors and the weight of the test point with respect to the damage factor.

[0045] Specifically, the maximum detection distance refers to the maximum distance from the test point that a group of sensors can detect to the axis of symmetry of the two receiving sensors in the same group of dual-element sensors.

[0046] Specifically, the similarity determination module is used for:

[0047] The signals received by the two receiving sensors in each group of sensors within the time window are determined and denoted as f. k1 (t) and f k2 (t), where f k1 (t) represents the signal received by one of the receiving sensors in the k-th group of sensors, f k2 (t) represents the signal received by another receiving sensor in the k-th sensor group, where k represents the sensor group number and t represents time;

[0048] The similarity of the signals received by the two receiving sensors in each group of sensors is determined by the following formula:

[0049]

[0050] Where, r k The similarity between the signals received by the two receiving sensors in the k-th group of sensors is represented by T, which represents the length of the time window.

[0051] Specifically, the damage factor determination module is used for:

[0052] The damage factor for each group of sensors is determined using the following formula:

[0053] DI k =1-r k

[0054] Where k is the sensor group number, DI k r represents the damage factor of the k-th sensor group. k This represents the similarity between the signals received by the two receiving sensors in the k-th group of sensors.

[0055] Specifically, the damage factor weight determination module is used for:

[0056] The ratio of the sum of the distances from the point to be measured to the midpoints of the excitation sensor and the two receiving sensors in each group of sensors to the distance from the excitation sensor to the midpoints of the two receiving sensors is determined by the following formula:

[0057]

[0058] Where k is the sensor group number, d k Let (x, y) be the distance ratio of the k-th group of sensors, and (x, y) be the coordinates of the point to be measured. ka y ka (x) represents the coordinates of the excitation sensor in the k-th sensor group. ks y ks ) represents the coordinates of the midpoint between the two receiving sensors in the k-th sensor group;

[0059] The weight of the test point with respect to the damage factor is determined by the following formula:

[0060]

[0061] Among them, R k (x,y) represents the weight of the test point (x,y) with respect to the damage factor of the k-th sensor group, γ k γ represents the parameter indicating the region affected by the damage factor of the preset k-th sensor group. k The larger the value, the larger the area affected by the damage factor of the k-th sensor group.

[0062] Specifically, the damage probability determination module is used for:

[0063] The damage probability of the test point is obtained by fusing the damage probabilities of the test point with respect to at least two sets of sensors using the following formula:

[0064]

[0065] Where I(x,y) is the damage probability of the test point (x,y), k is the sensor group number, N is the number of sensor groups participating in the fusion, and DI k R represents the damage factor of the k-th sensor group. k (x,y) represents the weight of the test point (x,y) with respect to the damage factor of the k-th sensor group, DI k R k (x,y) represents the damage probability of the test point (x,y) with respect to the k-th sensor group.

[0066] Optional, also includes:

[0067] The defect location determination module is used to determine the location of the defect as the test point where the damage probability exceeds a preset threshold.

[0068] Optional, also includes:

[0069] Based on the damage probability of each point on the top of the storage tank, draw a contour map of the damage probability of the top of the storage tank.

[0070] The damage status of the tank top is determined based on the damage probability contour map.

[0071] The beneficial effects of the above-described technical solutions provided in the embodiments of the present invention include at least the following:

[0072] By deploying dual-element sensors, the effective detection range of a single sensor group is expanded, improving detection accuracy. Because the effective detection range of a single sensor group is expanded, the number of sensor groups deployed is reduced, lowering detection costs, alleviating the workload and difficulty of sensor deployment, and reducing interference from the coupling effect between sensors on the detection results. The level of detection precision can be adjusted by changing the damage probability threshold, improving applicability. Furthermore, by drawing damage probability contour maps, the damage status of the tank top can be visually reflected, facilitating widespread application.

[0073] Other features and advantages of the invention will be set forth in the description which follows, and will be apparent in part from the description, or may be learned by practicing the invention. The objects and other advantages of the invention may be realized and obtained by means of the structures particularly pointed out in the written description, claims, and drawings.

[0074] The technical solution of the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. Attached Figure Description

[0075] The accompanying drawings are provided to further illustrate the invention and form part of the specification. They are used in conjunction with embodiments of the invention to explain the invention and do not constitute a limitation thereof. In the drawings:

[0076] Figure 1 This is a flowchart of a method for detecting defects on the top of a storage tank according to an embodiment of the present invention;

[0077] Figure 2 This is a schematic diagram of the sensor group and detection radius in an embodiment of the present invention;

[0078] Figure 3 This is a schematic diagram of the signal propagation and reception of the sensor group in an embodiment of the present invention;

[0079] Figure 4 This is a schematic diagram illustrating the probability of damage detection by the sensor group in an embodiment of the present invention;

[0080] Figure 5 This is a schematic diagram of a tank top defect detection system according to an embodiment of the present invention. Detailed Implementation

[0081] Exemplary embodiments of the present disclosure will now be described in more detail with reference to the accompanying drawings. While exemplary embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure may be implemented in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the scope of the disclosure to those skilled in the art.

[0082] To address the problems existing in the prior art, embodiments of the present invention provide a method and system for detecting defects on the top of storage tanks.

[0083] Example 1

[0084] Embodiment 1 of the present invention provides a method for detecting defects on the top of a storage tank, the process of which is as follows: Figure 1 As shown, it includes the following steps:

[0085] Step S1: Arrange multiple sets of sensors on the top of the storage tank. Each set of sensors includes a dual-element sensor and an excitation sensor. The dual-element sensor includes two receiving sensors. The excitation sensor is located on the axis of symmetry of the two receiving sensors. The receiving sensor receives signals emitted from the same set of excitation sensors.

[0086] like Figure 2 As shown, the solid line between the two receiving sensors is the axis of symmetry between the two receiving sensors in each group of sensors, and the excitation sensor in each group of sensors is arranged on this axis of symmetry.

[0087] The tank top structure can be simplified to a large, thin-plate structure. When using symmetrically distributed sensor arrays, identical response signals will be obtained. However, if the tank top is in an unhealthy state, differences in the response signals will occur.

[0088] Step S2: Select the time window length based on the preset maximum detection distance. The maximum detection distance refers to the maximum distance from the test point that a group of sensors can detect to the axis of symmetry of the two receiving sensors in the same group of dual-element sensors. For example... Figure 2 As shown, the solid line between the two receiving sensors is the axis of symmetry between them. Assuming a defect exists, the distance from the defect to the axis of symmetry is the detection distance. Figure 2 The detection radius d in the figure, when the defect point is Figure 2 The distance from the defect point to the axis of symmetry is the maximum detection distance when the sensor group shown can detect the farthest point to be measured. If the maximum detection distance of each sensor group is the same, then the time window length is selected according to the preset maximum detection distance. The larger the maximum detection distance, the larger the time window length. The time window length depends on the maximum detection distance desired in the subsequent imaging process.

[0089] Step S3: Using the Pearson correlation coefficient method, determine the similarity of the signals received by the two receiving sensors in each group of sensors within the time window.

[0090] Since it is difficult to maintain perfect consistency in the coupling of symmetrical distributed sensors during actual detection, the Pearson coefficient is chosen as a characteristic quantity for the difference parameter between two corresponding waveforms to measure the linear correlation between the two variables. The Pearson coefficient has displacement invariance and scale invariance, and can be used to measure the relationship without considering the influence of the received waveform amplitude or the influence of small phase shifts during propagation.

[0091] Figure 3 The two receiving sensors R1 and R2 are symmetrically distributed. The waveguide exciter (i.e., the excitation sensor) is located on the vertical line between the two receiving sensors. The excitation sensor reaches the two receiving points R1 and R2 at the same distance, i.e., L1 = L2. L is the direct path distance from the excitation point to the defect point. d1 and d2 are the path distances of the sound wave after bypassing the defect and reaching R1 and R2 respectively.

[0092] Damage affects the waveguide propagation path, causing changes in the received signal of the dual-element sensor and resulting in differences between the received signals. Therefore, by adopting a symmetrical layout and analyzing the differences in the structural response signals of the symmetrical monitoring channels, damage information can be identified.

[0093] When a defect exists, the signals received by R1 and R2 are:

[0094]

[0095] In the formula, R1(t) represents the signal received by the sensor at R1, L(t) represents the signal propagating from the excitation point to the defect point, L1(t) represents the signal propagating from the excitation sensor to the receiving point R1, d1(t) represents the signal that the sound wave reaches the receiving sensor R1 directly after bypassing the defect point, R2(t) represents the signal received by the sensor at R2, L2(t) represents the signal propagating from the excitation sensor to the receiving point R2, d2(t) represents the signal that the sound wave reaches the receiving sensor R2 directly after bypassing the defect point, and t represents time.

[0096] In traditional reference-based detection models, the defect scattering signal received at each receiving point is obtained by subtracting the received waveform with defects from the reference signal without defects under the same conditions. Therefore, the defect scattering signal obtained by solving R1 and R2 is:

[0097]

[0098] Where D1(t) represents the defect scattering signal at receiving point R1, D2(t) represents the defect scattering signal at receiving point R2, and R 10 (t) represents the reference signal under the condition that the receiving point R1 is defect-free, R 20 (t) represents the reference signal when the receiving point R2 is defect-free.

[0099] Because of the symmetrical relationship between the receiving points, when a defect exists, the signals received by R1 and R2 along paths L1 and L2 are exactly the same, and equal to the signals received by R1 and R2 when there is no defect. That is:

[0100]

[0101] Based on equations 1 and 2, we further obtain:

[0102]

[0103] Substituting equation 3 into equation 4, we get:

[0104]

[0105] That is:

[0106] D1(t)-D2(t)=d1(t)-d2(t) Equation 6

[0107] Therefore, when the receiving points are symmetrically arranged, the difference between the signals of the two receiving sensors includes the scattered signals from the defect to the two detection positions. Equation 6 shows that the difference between the defect scattered signals received by the two receiving sensors is the difference between the signals of the defect scattered signals propagating to the two receiving sensors.

[0108] The method of determining the similarity of signals received by two receiving sensors in each group of sensors within the time window using the Pearson correlation coefficient includes the following steps:

[0109] The signals received by the two receiving sensors in each group of sensors within the time window are determined and denoted as f. k1 (t) and f k2 (t), where f k1 (t) represents the signal received by one of the receiving sensors in the k-th group of sensors, f k2 (t) represents the signal received by another receiving sensor in the k-th sensor group, where k represents the sensor group number and t represents time;

[0110] The similarity of the signals received by the two receiving sensors in each group of sensors is determined by the following formula:

[0111]

[0112] Where, r k The similarity between the signals received by the two receiving sensors in the k-th group of sensors is represented by T, which represents the length of the time window.

[0113] Step S4: Determine the damage factor of the corresponding sensor group based on the similarity of the signals received by the two receiving sensors in each group of sensors.

[0114] The correlation coefficient of the wave packet decomposed at corresponding positions of the two sets of response signals (i.e., R below) k (x,y) and DI k Calculate them separately.

[0115] The step of determining the impairment factor for a corresponding sensor group based on the similarity of the signals received by two receiving sensors in each group includes the following steps:

[0116] The damage factor for each group of sensors is determined using the following formula:

[0117] DI k =1-r k

[0118] Where k is the sensor group number, DI k r represents the damage factor of the k-th sensor group. k This represents the similarity between the signals received by the two receiving sensors in the k-th group of sensors.

[0119] Step S5: Determine the weight of the test point with respect to the damage factor based on the preset size of the area affected by the damage factor and the positional relationship between the test point and the corresponding sensor group.

[0120] The step of determining the weight of the test point with respect to the damage factor based on the preset size of the area affected by the damage factor and the positional relationship between the test point and the corresponding sensor group includes the following steps:

[0121] The ratio of the sum of the distances from the point to be measured to the midpoints of the excitation sensor and the two receiving sensors in each group of sensors to the distance from the excitation sensor to the midpoints of the two receiving sensors is determined by the following formula:

[0122]

[0123] Where k is the sensor group number, d k Let (x, y) be the distance ratio of the k-th group of sensors, and (x, y) be the coordinates of the point to be measured. ka y ka (x) represents the coordinates of the excitation sensor in the k-th sensor group. ks y ks ) represents the coordinates of the midpoint between the two receiving sensors in the k-th sensor group;

[0124] The weight of the test point with respect to the damage factor is determined by the following formula:

[0125]

[0126] Among them, R k(x,y) represents the weight of the test point (x,y) with respect to the damage factor of the k-th sensor group, γ k γ represents the parameter indicating the region affected by the damage factor of the preset k-th sensor group. k The larger the value of γ, the larger the area affected by the damage factor of the k-th sensor group. The value of γ can be adjusted according to actual detection needs. k The size of the area affected by the damage is controlled by the size of the area.

[0127] Step S6: Determine the damage probability of the test point based on the damage factor of each sensor group and the weight of the test point with respect to the damage factor.

[0128] like Figure 4 As shown, the damage probability distribution of the sensing path formed by a group of sensors is elliptical, with the foci of the ellipse being the corresponding excitation sensor and the receiving sensor. Since the two receiving sensors are symmetrically distributed, therefore... Figure 4 The midpoint of the two receiving sensors is considered as the position of the receiving sensors. The intensity of the color indicates the magnitude of the damage probability. The damage probability at the line connecting the foci of the ellipse is 0, and the damage probability outside the edge of the ellipse is also 0. This is because when there is a defect distribution at the central axis of the dual-element sensor, the received waveforms of R1 and R2 remain symmetrical as in the healthy state. As the defect detection radius increases, the difference in the defect modulation signal increases. However, when the detection radius increases to a certain distance, the impact of the difference in the response signal caused by the defect can be ignored. That is, the difference increases in the first range, and the difference decreases after passing a threshold.

[0129] like Figure 4 As shown, a single sensing path cannot determine the location of the damage. To determine the location of the damage, it is necessary to fuse the damage probability images of multiple sensing paths.

[0130] The step of determining the damage probability of a test point based on the damage factor of each sensor group and the weight of the test point with respect to the damage factor includes the following steps:

[0131] The damage probability of the test point is obtained by fusing the damage probabilities of the test point with respect to at least two sets of sensors using the following formula:

[0132]

[0133] Where I(x,y) is the damage probability of the test point (x,y), k is the sensor group number, N is the number of sensor groups participating in the fusion, and DI k R represents the damage factor of the k-th sensor group. k (x,y) represents the weight of the test point (x,y) with respect to the damage factor of the k-th sensor group, DI k R k(x,y) represents the damage probability of the test point (x,y) with respect to the k-th sensor group.

[0134] The larger I(x,y) is, the greater the probability that the test point (x,y) will be damaged.

[0135] Step S7: The test points with damage probabilities exceeding a preset threshold are identified as the locations of defects.

[0136] Step S8: Draw a contour map of the damage probability of the tank top based on the damage probability of each point on the tank top; determine the damage status of the tank top based on the contour map of the damage probability.

[0137] In the method described in this embodiment, by arranging dual-element sensors, the effective detection range of a single sensor group is expanded, thereby improving detection accuracy. Because the effective detection range of a single sensor group is expanded, the number of sensor groups required is reduced, lowering detection costs, alleviating the workload and difficulty of sensor arrangement, and reducing the interference of sensor coupling effects on detection results. By adjusting the damage probability threshold, the detection precision can be adjusted, improving applicability. Furthermore, by drawing a damage probability contour map, the damage status of the tank top can be visually reflected, facilitating widespread application.

[0138] Those skilled in the art can change the above order without departing from the scope of protection of this disclosure.

[0139] Example 2

[0140] Embodiment 2 of the present invention provides a tank top defect detection system, the structure of which is as follows: Figure 5 As shown, it includes:

[0141] The sensor group arrangement module 100 is used to arrange multiple groups of sensors on the top of the storage tank. Each group of sensors includes a dual-element sensor and an excitation sensor. The dual-element sensor includes two receiving sensors. The excitation sensor is located on the axis of symmetry of the two receiving sensors. The receiving sensor receives signals emitted from the same group of excitation sensors.

[0142] The time window selection module 200 is used to select the length of the time window based on the preset maximum detection distance;

[0143] The similarity determination module 300 is used to determine the similarity of the signals received by two receiving sensors in each group of sensors within the time window using the Pearson correlation coefficient method.

[0144] The damage factor determination module 400 is used to determine the damage factor of the corresponding sensor group based on the similarity of the signals received by the two receiving sensors in each group of sensors.

[0145] The damage factor weight determination module 500 is used to determine the weight of the test point with respect to the damage factor based on the size of the damage factor influence area and the positional relationship between the test point and the corresponding sensor group.

[0146] The damage probability determination module 600 is used to determine the damage probability of the test point based on the damage factor of each group of sensors and the weight of the test point with respect to the damage factor.

[0147] The maximum detection distance refers to the maximum distance from the test point that a group of sensors can detect to the axis of symmetry of the two receiving sensors in the same group of dual-element sensors.

[0148] The similarity determination module 300 is specifically used for:

[0149] The signals received by the two receiving sensors in each group of sensors within the time window are determined and denoted as f. k1 (t) and f k2 (t), where f k1 (t) represents the signal received by one of the receiving sensors in the k-th group of sensors, f k2 (t) represents the signal received by another receiving sensor in the k-th sensor group, where k represents the sensor group number and t represents time;

[0150] The similarity of the signals received by the two receiving sensors in each group of sensors is determined by the following formula:

[0151]

[0152] Where, r k The similarity between the signals received by the two receiving sensors in the k-th group of sensors is represented by T, which represents the length of the time window.

[0153] The damage factor determination module 400 is specifically used for:

[0154] The damage factor for each group of sensors is determined using the following formula:

[0155] DI k =1-r k

[0156] Where k is the sensor group number, DI k r represents the damage factor of the k-th sensor group. k This represents the similarity between the signals received by the two receiving sensors in the k-th group of sensors.

[0157] The damage factor weight determination module 500 is specifically used for:

[0158] The ratio of the sum of the distances from the point to be measured to the midpoints of the excitation sensor and the two receiving sensors in each group of sensors to the distance from the excitation sensor to the midpoints of the two receiving sensors is determined by the following formula:

[0159]

[0160] Where k is the sensor group number, d k Let (x, y) be the distance ratio of the k-th group of sensors, and (x, y) be the coordinates of the point to be measured. ka y ka (x) represents the coordinates of the excitation sensor in the k-th sensor group. ks y ks ) represents the coordinates of the midpoint between the two receiving sensors in the k-th sensor group;

[0161] The weight of the test point with respect to the damage factor is determined by the following formula:

[0162]

[0163] Among them, R k (x,y) represents the weight of the test point (x,y) with respect to the damage factor of the k-th sensor group, γ k γ represents the parameter indicating the region affected by the damage factor of the preset k-th sensor group. k The larger the value, the larger the area affected by the damage factor of the k-th sensor group.

[0164] The damage probability determination module 600 is specifically used for:

[0165] The damage probability of the test point is obtained by fusing the damage probabilities of the test point with respect to at least two sets of sensors using the following formula:

[0166]

[0167] Where I(x,y) is the damage probability of the test point (x,y), k is the sensor group number, N is the number of sensor groups participating in the fusion, and DI k R represents the damage factor of the k-th sensor group. k (x,y) represents the weight of the test point (x,y) with respect to the damage factor of the k-th sensor group, DI k R k (x,y) represents the damage probability of the test point (x,y) with respect to the k-th sensor group.

[0168] The tank top defect detection system provided in Embodiment 2 of the present invention further includes: a defect location determination module 700, used to determine the test point where the damage probability exceeds a preset threshold as the location of the defect.

[0169] The tank top defect detection system provided in Embodiment 2 of the present invention further includes: a damage probability contour map drawing module 800, used to draw a damage probability contour map of the tank top based on the damage probability of each point on the tank top; and to determine the damage status of the tank top based on the damage probability contour map.

[0170] In this embodiment, by arranging dual-element sensors, the effective detection range of a single sensor group is expanded, and the detection accuracy is improved. Because the effective detection range of a single sensor group is expanded, the number of sensor groups arranged is reduced, the detection cost is lowered, the workload and difficulty of arranging sensors are reduced, and the interference of the coupling effect between sensors on the detection results is reduced. By adjusting the damage probability threshold, the detection precision can be adjusted, thus improving applicability. By drawing a damage probability contour map, the damage status of the tank top can be intuitively reflected, which is convenient for promotion and application.

[0171] Regarding the system in the above embodiments, the specific ways in which each module performs operations have been described in detail in the embodiments related to the method, and will not be elaborated here.

[0172] Any modifications, additions, and equivalent substitutions made within the scope of the principles of this invention shall still fall within the patent coverage of this invention.

Claims

1. A method for detecting defects on the top of a storage tank, characterized in that, Includes the following steps: Multiple sets of sensors are arranged on the top of the storage tank. Each set of sensors includes a dual-element sensor and an excitation sensor. The dual-element sensor includes two receiving sensors. The excitation sensor is located on the axis of symmetry of the two receiving sensors. The receiving sensor receives signals emitted from the same set of excitation sensors. Select the time window length based on the preset maximum detection distance; Using the Pearson correlation coefficient method, the similarity of the signals received by two receiving sensors in each group of sensors within the time window is determined respectively; The damage factor of the corresponding sensor group is determined based on the similarity of the signals received by the two receiving sensors in each group. Based on the size of the influence area of ​​the damage factor and the positional relationship between the test point and the corresponding sensor group, the weight of the test point with respect to the damage factor is determined. The damage probability of the test point is determined based on the damage factor of each sensor group and the weight of the test point with respect to the damage factor. The step of determining the impairment factor for a corresponding sensor group based on the similarity of the signals received by two receiving sensors in each group includes the following steps: The damage factor for each group of sensors is determined using the following formula: Where k is the sensor group number. Let r represent the damage factor of the k-th sensor group. k This represents the similarity between the signals received by two receiving sensors in the k-th group of sensors; The step of determining the weight of the test point with respect to the damage factor based on the preset size of the area affected by the damage factor and the positional relationship between the test point and the corresponding sensor group includes the following steps: The ratio of the sum of the distances from the point to be measured to the midpoints of the excitation sensor and the two receiving sensors in each group of sensors to the distance from the excitation sensor to the midpoints of the two receiving sensors is determined by the following formula: Where k is the sensor group number, d k Let (x, y) be the distance ratio of the k-th group of sensors, and (x, y) be the coordinates of the point to be measured. ka y ka (x) represents the coordinates of the excitation sensor in the k-th group of sensors. ks y ks ) represents the coordinates of the midpoint between two receiving sensors in the k-th sensor group; The weight of the test point with respect to the damage factor is determined by the following formula: in, This represents the weight of the damage factor of the test point (x, y) with respect to the k-th sensor group. The parameter represents the region affected by the damage factor of the preset k-th sensor group. The larger the value, the larger the area affected by the damage factor of the k-th sensor group; The step of determining the damage probability of a test point based on the damage factor of each sensor group and the weight of the test point with respect to the damage factor includes the following steps: The damage probability of the test point is obtained by fusing the damage probabilities of the test point with respect to at least two sets of sensors using the following formula: in, Let k be the damage probability of the test point (x, y), k be the sensor group number, and N be the number of sensor groups participating in the fusion. This represents the damage factor of the k-th sensor group. This represents the weight of the damage factor of the test point (x, y) with respect to the k-th sensor group. This represents the damage probability of the test point (x, y) with respect to the k-th sensor group.

2. The method as described in claim 1, characterized in that, The maximum detection distance refers to the maximum distance from the test point that a group of sensors can detect to the axis of symmetry of the two receiving sensors in the same group of dual-element sensors.

3. The method as described in claim 1, characterized in that, The method of determining the similarity of signals received by two receiving sensors in each group of sensors within the time window using the Pearson correlation coefficient includes the following steps: The signals received by the two receiving sensors in each group of sensors within the time window are determined and denoted as f. k1 (t) and f k2 (t), where f k1 (t) represents the signal received by one of the receiving sensors in the k-th group of sensors, f k2 (t) represents the signal received by another receiving sensor in the k-th sensor group, where k represents the sensor group number and t represents time; The similarity of the signals received by the two receiving sensors in each group of sensors is determined by the following formula: Where, r k The similarity between the signals received by the two receiving sensors in the k-th group of sensors is represented by T, which represents the length of the time window.

4. The method as described in claim 1, characterized in that, It also includes the following steps: The test points whose damage probability exceeds a preset threshold are identified as the locations of the defects.

5. The method according to any one of claims 1-4, characterized in that, It also includes the following steps: Based on the damage probability of each point on the top of the storage tank, draw a contour map of the damage probability of the top of the storage tank. The damage status of the tank top is determined based on the damage probability contour map.

6. A tank top defect detection system, characterized in that, include: A sensor group arrangement module is used to arrange multiple groups of sensors on the top of a storage tank. Each group of sensors includes a dual-element sensor and an excitation sensor. The dual-element sensor includes two receiving sensors. The excitation sensor is located on the axis of symmetry of the two receiving sensors. The receiving sensor receives signals emitted from the same group of excitation sensors. The time window selection module is used to select the length of the time window based on the preset maximum detection distance; The similarity determination module is used to determine the similarity of signals received by two receiving sensors in each group of sensors within the time window using the Pearson correlation coefficient method. The damage factor determination module is used to determine the damage factor of the corresponding sensor group based on the similarity of the signals received by the two receiving sensors in each group of sensors. The damage factor weight determination module is used to determine the weight of the test point with respect to the damage factor based on the size of the preset damage factor influence area and the positional relationship between the test point and the corresponding sensor group. The damage probability determination module is used to determine the damage probability of the test point based on the damage factor of each group of sensors and the weight of the test point with respect to the damage factor. The damage factor determination module is specifically used for: The damage factor for each group of sensors is determined using the following formula: Where k is the sensor group number. Let r represent the damage factor of the k-th sensor group. k This represents the similarity between the signals received by two receiving sensors in the k-th group of sensors; The damage factor weight determination module is specifically used for: The ratio of the sum of the distances from the point to be measured to the midpoints of the excitation sensor and the two receiving sensors in each group of sensors to the distance from the excitation sensor to the midpoints of the two receiving sensors is determined by the following formula: Where k is the sensor group number, d k Let (x, y) be the distance ratio of the k-th group of sensors, and (x, y) be the coordinates of the point to be measured. ka y ka (x) represents the coordinates of the excitation sensor in the k-th group of sensors. ks y ks ) represents the coordinates of the midpoint between two receiving sensors in the k-th sensor group; The weight of the test point with respect to the damage factor is determined by the following formula: in, This represents the weight of the damage factor of the test point (x, y) with respect to the k-th sensor group. The parameter represents the region affected by the damage factor of the preset k-th sensor group. The larger the value, the larger the area affected by the damage factor of the k-th sensor group; The damage probability determination module is specifically used for: The damage probability of the test point is obtained by fusing the damage probabilities of the test point with respect to at least two sets of sensors using the following formula: in, Let k be the damage probability of the test point (x, y), k be the sensor group number, and N be the number of sensor groups participating in the fusion. This represents the damage factor of the k-th sensor group. This represents the weight of the damage factor of the test point (x, y) with respect to the k-th sensor group. This represents the damage probability of the test point (x, y) with respect to the k-th sensor group.

7. The system as described in claim 6, characterized in that, The maximum detection distance refers to the maximum distance from the test point that a group of sensors can detect to the axis of symmetry of the two receiving sensors in the same group of dual-element sensors.

8. The system as described in claim 6, characterized in that, The similarity determination module is specifically used for: The signals received by the two receiving sensors in each group of sensors within the time window are determined and denoted as f. k1 (t) and f k2 (t), where f k1 (t) represents the signal received by one of the receiving sensors in the k-th group of sensors, f k2 (t) represents the signal received by another receiving sensor in the k-th sensor group, where k represents the sensor group number and t represents time; The similarity of the signals received by the two receiving sensors in each group of sensors is determined by the following formula: Where, r k The similarity between the signals received by the two receiving sensors in the k-th group of sensors is represented by T, which represents the length of the time window.

9. The system as described in claim 6, characterized in that, Also includes: The defect location determination module is used to determine the location of the defect as the test point where the damage probability exceeds a preset threshold.

10. The system as described in any one of claims 6-9, characterized in that, Also includes: Based on the damage probability of each point on the top of the storage tank, draw a contour map of the damage probability of the top of the storage tank. The damage status of the tank top is determined based on the damage probability contour map.

Images