Edge computing based receiving station multi-source data fusion method

Abstract

The application discloses a receiving station multi-source data fusion method based on edge calculation and relates to the technical field of edge calculation, and comprises the following steps: generating a ring direction position of a ring gap, constructing a non-projected area based on three-dimensional space coordinate data and the ring direction position of the ring gap; dividing a tank top plane of a tank top structure into a plurality of ring sectors according to the non-projected area; performing edge side distributed collaborative convergence operation on the ring sectors to obtain a reference quantity set; calculating an index value of the ring sector according to the reference quantity set, and screening out a target sector according to the index value and a symbol comparison result. The application realizes quantitative expression of the strength of the sector heat influence, thereby providing clear indication for fixed-point inspection.

Description

Technical Field

[0001] This invention relates to the field of edge computing technology, and in particular to a method for multi-source data fusion at receiving stations based on edge computing. Background Technology

[0002] During long-term operation, liquefied natural gas (LNG) receiving terminals require continuous monitoring of large, full-containment LNG storage tanks to ensure stable delivery of evaporating gas in the upper space and the safety of the tank top structure. The tank top area is typically equipped with temperature sensors, strain monitoring devices, and flow or pressure measuring points related to evaporating gas extraction. These monitoring devices are distributed in different circumferential locations within the tank to reflect the overall operational status of the tank top in terms of temperature changes, structural deformation, and evaporating gas distribution. With the increasing prevalence of edge computing capabilities in industrial settings, receiving terminals are deployed near the tank area, enabling data to be collected, processed, and analyzed close to the field, reducing data transmission burden and improving the timeliness of analysis results. Due to the large spatial area and complex operating conditions of the tank top structure, operational differences may arise in different circumferential locations due to factors such as aging of the insulation layer, moisture absorption, material fatigue, and changes in evaporating gas extraction conditions. These differences often manifest as localized variations in temperature, strain, and evaporating gas volume. Failure to identify these localized anomalies can increase the uncertainty in tank operation and management.

[0003] In existing technologies, multi-source data processing is typically conducted on a tank-by-tank basis. This includes averaging temperature data from all sensors, performing uniform processing on strain data from all locations, or calculating the overall change in evaporated gas volume. This approach ignores spatial differences along the circumferential direction of the tank top, meaning that localized temperature increases, structural deformations, or abnormal evaporated gas volumes are often masked by the smoothing effect of the overall data, making it difficult to promptly identify localized risks that could affect tank operation. Summary of the Invention

[0004] The purpose of this invention is to address the shortcomings of existing technologies that ignore the spatial differences in the circumferential position of the tank top, and to propose a multi-source data fusion method for receiving stations based on edge computing.

[0005] To address the problems existing in the prior art, the present invention adopts the following technical solution: A multi-source data fusion method for receiving stations based on edge computing includes: S1. Generate the circumferential position of the annular gap, and construct a non-projected region based on the three-dimensional spatial coordinate data and circumferential position of the annular gap; S2. Based on the non-projection area, the tank top plane of the tank top structure is divided into multiple circumferential sectors; S3. Perform edge-side distributed collaborative aggregation operation on the circumferential sector to obtain the baseline quantity set; S4. Calculate the index value of the circumferential sector based on the benchmark set, and select the target sector based on the index value and the sign comparison result. S5. Perform point-to-point checks on the target sector based on the data fusion value of the target sector.

[0006] Preferably, the circumferential location for generating the annular gap includes: The top structure of the liquefied natural gas storage tank in the positioning receiving station; The annular gap between the suspension plate in the top structure of the positioning tank and the inner tank wall of the liquefied natural gas storage tank; Obtain the three-dimensional spatial coordinate data of the annular gap; Angle calculations are performed on the three-dimensional spatial coordinate data of the annular gap to obtain the circumferential position of the annular gap.

[0007] Preferably, a non-projected region is constructed based on the three-dimensional spatial coordinate data and circumferential position of the annular gap, including: The radial distance of the annular gap is calculated by analyzing its three-dimensional spatial coordinate data to obtain its radial position. Obtain the three-dimensional geometric data of the tank top structure; Based on the circumferential and radial positions, each three-dimensional geometric point in the three-dimensional geometric data is determined and marked to obtain the gap point; The gap points are aggregated to form a non-projected region.

[0008] Preferably, the tank top plane of the tank top structure is divided into multiple circumferential sectors based on the non-projection area, including: Geometric constraint processing is performed on the three-dimensional geometric data of the tank top structure based on the non-projection region to determine the reference projection range; Within the reference projection range, projection calculations are performed on the three-dimensional geometric data of the tank top structure to obtain two-dimensional projection data; The two-dimensional projection data is reconstructed into a plane to obtain the tank top plane. The top plane of the tank is divided into multiple circumferential sectors.

[0009] Preferably, an edge-side distributed collaborative aggregation operation is performed on the circumferential sector to obtain a set of reference quantities, including: Deploy edge acquisition terminals in each circumferential sector; Temperature value sequences for each circumferential sector are collected using an edge acquisition terminal; The temperature value sequences of each circumferential sector are locally aggregated, and all temperature value sequences are aggregated to obtain the circumferential average temperature of the tank top structure. The strain value sequence of the circumferential sector is acquired through the edge acquisition terminal; The strain value sequences of each circumferential sector are locally aggregated, and all strain value sequences are aggregated to obtain the circumferential average strain of the tank top structure. The evaporation gas indicator value sequence of each circumferential sector is collected through the edge acquisition terminal; The evaporation gas indicator value sequences of each circumferential sector are locally aggregated, and all evaporation gas indicator value sequences are aggregated to obtain the uniform distribution value of evaporation gas in the tank top structure. The circumferential average temperature, circumferential average strain, and uniform distribution of evaporated gas are collected as the reference values ​​for the circumferential sector.

[0010] Preferably, the index value of the circumferential sector is calculated based on the reference quantity set, including: The temperature values ​​of the circumferential sectors are calculated from the temperature value sequence of the circumferential sectors; The temperature deviation is obtained by subtracting the average circumferential temperature value from the reference set from the temperature value of the circumferential sector. The strain values ​​of the circumferential sector are calculated from the strain value sequence of the circumferential sector; The strain deviation is obtained by subtracting the strain value of the circumferential sector from the average strain value in the reference set. The evaporation gas indicator values ​​of the circumferential sector are calculated from the evaporation gas indicator value sequence of the circumferential sector; The evaporation gas indication deviation is obtained by subtracting the evaporation gas uniform distribution value in the reference quantity set from the evaporation gas indication value of the circumferential sector. The sign of the evaporating gas indication deviation is extracted to obtain the deviation sign value; The absolute values ​​of temperature deviation, strain deviation, and deviation sign are multiplied sequentially to obtain the index values ​​of the circumferential sector.

[0011] Preferably, the formula for calculating the deviation of the evaporating gas indication is as follows: ; In the formula, It is an evaporation gas indication deviation. This is the value of the evaporated gas indicator. It is the value of uniform distribution of evaporated gas.

[0012] Preferably, the target sector is selected based on the index value and the sign comparison result, including: The average index value is obtained by averaging the index values ​​of all circumferential sectors. The difference between the index value and the average index value of the circumferential sector is calculated to obtain the index difference. Sign comparison was performed on temperature deviation and strain deviation to obtain the sign comparison results; Based on the index difference and sign comparison results, the target sector is selected from all circumferential sectors.

[0013] Preferably, the targeted sector is checked based on the data fusion value of the target sector, including: Data fusion is performed on temperature deviation and strain deviation to obtain fused data values; The data fusion value is defined as the additional heat index value of the target sector; The target sector is inspected based on the additional heat index value.

[0014] Preferably, the formula for calculating the data fusion value is as follows: ; In the formula, It's a temperature deviation. It is strain deviation. , These are the coefficients of a linear combination. It is the data fusion value.

[0015] Compared with the prior art, the beneficial effects of the present invention are: 1. This invention measures the annular gaps in the tank top structure using three-dimensional coordinates and calculates their circumferential positions. Combined with the non-projection area formed by the gaps, the invention applies geometric constraints to the tank top structure. This ensures that subsequent projection and partitioning operations are based solely on the effective structural area and are not affected by structural obstructions or discontinuous gaps. By accurately reconstructing the tank top plane and dividing it into multiple circumferential sectors, each sector corresponds to a clearly defined and stable actual area in the tank top space. This guarantees the consistency of spatial attribution of multi-source data, thereby achieving a fine characterization of the circumferential positional differences of the tank top structure. This provides a clear and reliable spatial basis for subsequent zone monitoring, zone calculation, and anomaly localization.

[0016] 2. This invention decentralizes data processing tasks to edge acquisition terminals corresponding to each circumferential sector on the tank top, completing local aggregation and inter-sectoral collaborative aggregation near the acquisition source. This reduces large-scale remote backhaul of raw data, significantly shortening the computation link. Through local processing and collaborative aggregation of multi-source data at the edge, a set of benchmark quantities consisting of the circumferential average temperature, the circumferential average strain, and the uniform distribution value of evaporated gas can be formed in real time. This allows for the immediate quantification of the overall operating status of the tank top structure, improving data processing efficiency and overall response speed. Clear deviation indications can be generated at the early stage of abnormalities in the tank top status, thereby enhancing the timeliness and reliability of operation monitoring.

[0017] 3. This invention calculates the temperature deviation, strain deviation, and evaporation gas indication deviation of each circumferential sector, generates sector index values ​​by combining the relationship between the deviation amplitude and the deviation direction, and then filters target sectors based on the index difference and sign comparison results. This allows spatial areas showing significant changes in the tank top structure to be accurately identified from numerous sectors. Furthermore, it constructs a data fusion value through temperature deviation and strain deviation and defines it as an additional heat index value, realizing a quantitative expression of the strength of the thermal influence of the sector, thereby providing clear indications for fixed-point inspection. Attached Figure Description

[0018] The accompanying drawings, which are included to provide a further understanding of the invention and form part of this application, illustrate exemplary embodiments of the invention and, together with their description, serve to explain the invention and do not constitute an undue limitation thereof. In the drawings: Figure 1 This is a flowchart illustrating a multi-source data fusion method for receiving stations based on edge computing, provided in an embodiment of the present invention. Detailed Implementation

[0019] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments.

[0020] Example: This example provides a multi-source data fusion method for receiving stations based on edge computing. See [link to example]. Figure 1 Specifically, including: S1. Generate the circumferential position of the annular gap, and construct a non-projected region based on the three-dimensional spatial coordinate data and circumferential position of the annular gap; In an embodiment of the present invention, the circumferential position for generating the annular gap includes: The top structure of the liquefied natural gas storage tank in the positioning receiving station; Specifically, the physical location of the target liquefied natural gas storage tank is first determined within the receiving station. A laser rangefinder is used to measure the vertical height of the tank, determining the vertical distance from the bottom foundation surface to the outermost top component, thus defining the overall vertical range of the tank. Next, a total station is used to scan the horizontal projection boundary of the tank's top area, obtaining the planar coordinates of each endpoint of the outermost edge of the tank top. Connecting these endpoints forms a closed profile, thus determining the horizontal distribution range of the tank top structure. Finally, through manual inspection combined with tape measure measurement, the installation positions of the insulation structure and auxiliary components supported by the tank top structure are identified, determining the connection points between these components and the main tank top structure, thereby clarifying the complete structural system range of the tank top structure and completing the positioning of the tank top structure.

[0021] The annular gap between the suspension plate in the top structure of the positioning tank and the inner tank wall of the liquefied natural gas storage tank; Specifically, the process begins by entering the tank top structure and using strong lighting equipment to observe and identify the plate-like components suspended below the tank top structure via connecting members. By measuring the planar dimensions of these plate-like components and the location of their suspension connection points with the tank top structure, the component is identified as a suspension plate, and the outer edge of the suspension plate is determined. Next, the inner diameter and vertical extension height of the vertical enclosure structure are measured to identify it as the inner tank wall of the liquefied natural gas storage tank, thus determining the distribution area of ​​the inner tank wall within the tank top structure. Finally, the distance between the outer edge of the suspension plate and the inner tank wall is measured segment by segment along the circumference of the storage tank. When the measured distance value is greater than zero and is continuously or intermittently distributed along the circumference, the area corresponding to this distance is identified as an annular gap, thus completing the location of the annular gap.

[0022] Specifically, the tank top structure refers to the structural system located on top of the liquefied natural gas (LNG) storage tank within the LNG receiving terminal. It serves to support the tank top insulation structure, related auxiliary components, and to form an isolation space between the tank top structure and the external structure. The suspension plate is a plate-shaped component installed inside the tank top structure. It is suspended below the tank top structure by connecting components to support the tank top insulation material and form a stable load-bearing plane inside the tank top. The inner tank wall of the LNG storage tank is a vertical enclosure structure that constitutes the main body of the inner tank of the LNG storage tank. It is used to limit the storage space of LNG and bear the load generated by LNG.

[0023] Specifically, the annular gap refers to a continuous or discontinuous annular gap region formed in the circumferential direction between the outer edge of the suspension plate and the inner tank wall of the liquefied natural gas storage tank inside the top structure of the liquefied natural gas receiving station. This annular gap region is distributed along the circumference of the storage tank, and its formation is due to factors including the structural installation spacing between the suspension plate and the inner tank wall, the splicing relationship of components, and the layout requirements of the tank top insulation structure. As an objectively existing geometric transition region in the tank top structure, this annular gap is used to reflect the spatial relationship between the suspension plate and the inner tank wall, and is identified as an independent structural region in the geometric analysis and data processing of the tank top structure.

[0024] Obtain the three-dimensional spatial coordinate data of the annular gap; Specifically, a three-dimensional rectangular coordinate system is first established with the center of the bottom of the storage tank as the origin. The X and Y axes are located on the plane of the base surface of the bottom of the storage tank, and the Z axis extends vertically along the storage tank. Then, using a three-dimensional coordinate measuring instrument, at least twenty evenly distributed feature points are selected along the extension trajectory of the annular gap, including the inner and outer edge points of the annular gap. The inner edge points are points on the outer edge of the suspension plate, and the outer edge points are points on the corresponding positions of the inner tank wall. The X, Y, and Z coordinate values ​​of each feature point in the above-mentioned three-dimensional rectangular coordinate system are measured in sequence. Finally, the three-dimensional coordinate values ​​of all feature points are arranged in the circumferential distribution order of the annular gap to form a three-dimensional spatial coordinate dataset of the annular gap, thus completing the acquisition of three-dimensional spatial coordinate data.

[0025] Angle calculations are performed on the three-dimensional spatial coordinate data of the annular gap to obtain the circumferential position of the annular gap.

[0026] Specifically, maintaining the aforementioned three-dimensional rectangular coordinate system with the center of the tank bottom as the origin, a first horizontal direction defined along the tank bottom base surface, a second horizontal direction defined along the tank bottom base surface and perpendicular to the first horizontal direction, and a vertical direction defined extending vertically along the tank, the coordinate values ​​of all feature points in the annular gap's three-dimensional spatial coordinate data are extracted in the first and second horizontal directions respectively. Using the positive direction of the first horizontal direction as the starting reference for the polar angle, and the counterclockwise direction as the direction of polar angle increment, the arctangent function is used to calculate the initial angle between each feature point and the line connecting it to the origin relative to the positive direction of the first horizontal direction. When the coordinate value of the feature point in the first horizontal direction is greater than 0 and the coordinate value in the second horizontal direction is greater than 0, the initial angle is directly used as the polar angle value. When the coordinate value of the feature point in the first horizontal direction is less than 0 and the coordinate value in the second horizontal direction is greater than or equal to 0, π radians are added to the initial angle to obtain the polar angle value. When the coordinate value of the feature point in the first horizontal direction is less than 0 and the coordinate value in the second horizontal direction is less than 0, the initial angle is calculated based on the initial angle. The polar angle value is obtained by superimposing π radians on the base. When the coordinate value of the feature point in the first horizontal direction is equal to 0 and the coordinate value in the second horizontal direction is greater than 0, the polar angle value is set to π divided by 2 radians. When the coordinate value of the feature point in the first horizontal direction is equal to 0 and the coordinate value in the second horizontal direction is less than 0, the polar angle value is set to 3π divided by 2 radians. The polar angle values ​​of all feature points are standardized to the range of 0 to 2π radians. Then, all polar angle values ​​are sorted in ascending order. The minimum and maximum values ​​of the sorted polar angle values ​​are selected and the difference between the two is calculated. If the difference is less than or equal to 2π radians, the minimum value is used as the starting value and the maximum value is used as the ending value to form a single angle interval. If the difference is greater than 2π radians, the starting reference of the annular gap across the positive direction of the first horizontal direction is determined. The angle interval is divided into two sub-intervals. One sub-interval starts with the minimum value and ends with 2π radians, and the other sub-interval starts with 0 radians and ends with the maximum value. This single angle interval or two sub-intervals are the circumferential position of the annular gap.

[0027] Specifically, the three-dimensional spatial coordinate data of the annular gap refers to the coordinate information used to describe the position and shape of the annular gap in three-dimensional space, which reflects the spatial distribution of the annular gap inside the storage tank; the circumferential position refers to the angle range calculated based on the three-dimensional spatial coordinate data, which characterizes the circumferential distribution position of the annular gap along the storage tank, thereby reflecting the relative positional relationship of the annular gap in the circumferential direction of the storage tank.

[0028] In embodiments of the present invention, a non-projected region is constructed based on the three-dimensional spatial coordinate data and circumferential position of the annular gap, including: The radial distance of the annular gap is calculated by analyzing its three-dimensional spatial coordinate data to obtain its radial position. Specifically, a 3D laser scanner is used to scan the top structure of the liquefied natural gas storage tank. During scanning, the scanner is placed at multiple different locations around the top structure, ensuring that the scanning range of each location covers at least one-third of the top structure, and that there is no less than 10% overlap between the scanning ranges of adjacent locations. During the scanning process, data is collected one by one from the surface of the insulation layer, the surface of the suspension plate, the surface of the connecting components, and the top surface of the inner tank wall of the top structure to obtain raw point cloud data. Then, a statistical filtering method is used to process the raw point cloud data to remove noise points that are more than a preset threshold away from the surrounding point clouds. Then, a voxel grid filtering method is used to downsample the denoised point cloud data to make the point cloud density uniform and retain the geometric contour features of the top structure. The final set containing multiple spatial coordinate points is the 3D geometric data of the top structure.

[0029] Obtain the three-dimensional geometric data of the tank top structure; Specifically, within the defined tank space, spatial geometry is collected for each component of the tank top structure. The spatial position data of the surface of the tank top structure and its internal components are recorded point by point. The collected spatial position data is then organized into a geometric data set consisting of multiple three-dimensional geometric points. Each three-dimensional geometric point is used to describe the local geometric shape of the tank top structure in space. This three-dimensional geometric data fully reflects the spatial distribution of the tank top structure inside the tank, providing basic geometric information for subsequent spatial position-based judgment and marking.

[0030] Based on the circumferential and radial positions, each three-dimensional geometric point in the three-dimensional geometric data is determined and marked to obtain the gap point; Specifically, maintaining a three-dimensional rectangular coordinate system with the center of the tank bottom as the origin, a first horizontal direction defined along the tank bottom foundation surface, a second horizontal direction defined along the tank bottom foundation surface and perpendicular to the first horizontal direction, and a vertical direction defined extending vertically along the tank, each three-dimensional geometric point in the three-dimensional geometric data of the tank top structure is extracted. The coordinate values ​​of this point in the first and second horizontal directions are obtained respectively. The sum of the squares of the coordinate values ​​in the first and second horizontal directions is calculated, and the square root of this sum is taken to obtain the radial distance of the point. Simultaneously, with the positive direction of the first horizontal direction as the starting reference for the polar angle and the counterclockwise direction as the direction of polar angle increase, the arctangent function is used to calculate the... The initial angle corresponding to the line connecting the point and the origin is normalized to a polar angle range of 0 to 2π radians, taking into account the sign of the point's coordinates in the first and second horizontal directions. Then, the radial distance of the point is compared with the radial position interval of the annular gap to determine whether it is between the start and end values ​​of the interval. At the same time, the polar angle of the point is compared with the circumferential position interval of the annular gap. If the circumferential position of the annular gap is a single interval, it is determined whether the polar angle is within that interval. If it is two sub-intervals, it is determined whether the polar angle is within either sub-interval. If the radial distance of the point is within the radial position interval and the polar angle is within the circumferential position interval, then the three-dimensional geometric point is marked as the gap point.

[0031] The gap points are aggregated to form a non-projected region.

[0032] Specifically, all 3D geometric points marked as gap points are collected and arranged in ascending order of their corresponding polar angle values. A spatial distance threshold between adjacent gap points is set, which is 1.5 times the average point spacing corresponding to the point cloud density in the 3D geometric data of the tank top structure. A region growing algorithm is used to process the arranged gap points. The first gap point is used as the initial seed point, and the remaining gap points are traversed. The spatial straight-line distance between the current seed point and the remaining gap points is calculated. Gap points with a distance less than the set threshold are assigned to the sub-region where the current seed point is located, and these points are used as new seed points to continue traversing until no points that meet the conditions can be added. Then, the next gap point that has not been assigned to any sub-region is selected as the new initial seed point and the above operation is repeated. After the sub-region division of all gap points is completed, the gap points contained in all sub-regions are merged to form a complete set of gap points. The spatial range corresponding to this set of points is the non-projection region.

[0033] Specifically, the radial position of the annular gap refers to the spatial position range obtained by calculating the distance between the three-dimensional spatial coordinate points constituting the annular gap and the geometric center of the liquefied natural gas storage tank. It is used to characterize the distribution range of the annular gap in the outward direction from the center of the storage tank. The three-dimensional geometric data of the tank top structure refers to the set of spatial data used to describe the position and shape of each component of the tank top structure in three-dimensional space. This data is composed of multiple three-dimensional geometric points and is used to reflect the actual spatial outline of the tank top structure. The three-dimensional geometric point refers to the basic spatial unit constituting the three-dimensional geometric data. Each three-dimensional geometric point corresponds to the spatial coordinates of a specific location in the tank top structure and is used to describe the spatial distribution of that location inside the storage tank.

[0034] Specifically, gap points refer to three-dimensional geometric points in the three-dimensional geometric data of the tank top structure that are identified as being located within the annular gap space after being judged by their circumferential and radial positions. These three-dimensional geometric points correspond to the annular gap region formed between the outer edge of the suspension plate and the inner wall of the liquefied natural gas storage tank, and are used to characterize the actual spatial distribution of the annular gap in the tank top structure. Non-projection regions refer to the spatial regions formed by the aggregation of all gap points. These spatial regions correspond to the geometric regions in the tank top structure that are not suitable for subsequent planar projection processing. They are used to exclude space during the subsequent projection and region division of the tank top structure, thereby ensuring that the projection results are generated only based on the effective geometric regions of the tank top structure.

[0035] S2. Based on the non-projection area, the tank top plane of the tank top structure is divided into multiple circumferential sectors; In an embodiment of the present invention, the tank top plane of the tank top structure is divided into multiple circumferential sectors based on the non-projection area, including: Geometric constraint processing is performed on the three-dimensional geometric data of the tank top structure based on the non-projection region to determine the reference projection range; Specifically, a three-dimensional rectangular coordinate system is established, with the center of the tank bottom as the origin, a first horizontal direction defined along the base surface of the tank bottom, a second horizontal direction defined along the base surface of the tank bottom and perpendicular to the first horizontal direction, and a vertical direction defined extending vertically along the tank. All three-dimensional geometric points of the tank top structure's three-dimensional geometric data are traversed, and each point is compared to the point set corresponding to a non-projected area. If a point belongs to a non-projected area, it is removed from the three-dimensional geometric data, and the remaining points form a valid three-dimensional geometric point set. Then, the coordinate values ​​of each point in the valid three-dimensional geometric point set are extracted in the first horizontal direction, and the maximum and minimum values ​​are determined to define the range of the first horizontal direction. The coordinate values ​​of each point in the second horizontal direction are extracted, and the maximum and minimum values ​​are determined to define the range of the second horizontal direction. The coordinate values ​​of each point in the vertical direction are extracted, and the maximum and minimum values ​​are determined to define the range of the vertical direction. The spatial range jointly defined by the first horizontal direction range, the second horizontal direction range, and the vertical direction range is determined as the reference projection range.

[0036] Within the reference projection range, projection calculations are performed on the three-dimensional geometric data of the tank top structure to obtain two-dimensional projection data; Specifically, within the defined reference projection range, the vertical direction is selected as the projection direction. For each point in the effective set of three-dimensional geometric points, its coordinates in the first horizontal direction are extracted as the first plane coordinates after two-dimensional projection, and its coordinates in the second horizontal direction are extracted as the second plane coordinates after two-dimensional projection. This process yields the two-dimensional coordinates corresponding to each point. After extracting the coordinates of all points, the obtained two-dimensional coordinates are deduplicated, i.e., duplicate two-dimensional coordinate values ​​are removed. The deduplicated two-dimensional coordinates are then arranged in ascending order of the first plane coordinate values. The set of arranged two-dimensional coordinates is the two-dimensional projection data.

[0037] The two-dimensional projection data is reconstructed into a plane to obtain the tank top plane. Specifically, firstly, all two-dimensional coordinate points in the two-dimensional projection data are extracted. The extreme values ​​of the first and second plane coordinate values ​​corresponding to each point are calculated to determine the smallest bounding rectangle that can contain all coordinate points. Using the center of this rectangle as a reference, all coordinate points are traversed and the nearest neighboring points are connected to form the initial two-dimensional contour line. Next, the closure degree of the initial contour line is calculated. If there are unclosed gaps in the contour line, the endpoints at both ends of the gap are selected and coordinate points with the same distribution trend as the coordinate points around the gap are added until the contour line is completely closed. Then, the area inside the closed contour line is filled. Any blank position inside the contour line is selected, and the distance between the position and the existing coordinate points around it is determined. If the distance is less than the average spacing of the points in the two-dimensional projection data, the corresponding coordinate point is added until the area inside the closed contour line is evenly covered by coordinate points. The final complete closed plane is the tank top plane.

[0038] Specifically, geometric constraint processing refers to a method of spatially limiting the three-dimensional geometric data of the tank top structure based on the identification of non-projection areas, in order to clarify the effective structural area allowed to participate in subsequent geometric calculations; the reference projection range refers to the spatial range of the tank top structure that is suitable for planar projection after geometric constraint processing; the two-dimensional projection data refers to the data result formed after mapping the three-dimensional geometric data of the tank top structure to two-dimensional space along a predetermined direction within the reference projection range, in order to express the geometric distribution of the tank top structure in the two-dimensional plane.

[0039] The top plane of the tank is divided into multiple circumferential sectors.

[0040] Specifically, the projection point of the tank bottom center onto the tank top plane is taken as the central origin of the tank top plane. A straight line is drawn from the central origin along the positive direction of the first plane coordinate value as the dividing baseline. The number of circumferential sectors is determined according to the zoning analysis requirements of the tank top structure. 360 degrees is calculated and divided by this number to obtain the central angle angle corresponding to a single sector. Starting from the dividing baseline, multiple dividing lines are drawn in a counterclockwise direction with the central origin as the vertex, with the angle between the dividing lines and the baseline equal to the central angle angle, until the dividing lines cover the 360-degree range. At the same time, at least one circular dividing line is drawn with the central origin as the center. The radius of the circular dividing line is taken as the midpoint of the radial length of the tank top plane, so that the circular dividing line divides the tank top plane into inner and outer parts radially. The dividing lines and circular dividing lines intertwine to divide the tank top plane into multiple regional units with both circumferential and radial ranges. Each regional unit is a circumferential sector.

[0041] Specifically, the tank top plane refers to the geometric plane formed after two-dimensional projection data is reconstructed, which is used to characterize the overall shape of the tank top structure in the sense of horizontal projection; the circumferential sector refers to multiple regional units formed by dividing the tank top plane along the circumferential and radial directions of the storage tank. Each circumferential sector corresponds to a circumferential range on the tank top plane, which is used for subsequent partition analysis and data processing of the tank top structure.

[0042] Specifically, the reason for dividing the tank top plane of the tank top structure into multiple circumferential sectors based on the non-projection area is that the non-projection area corresponding to the annular gap formed between the outer edge of the suspension plate and the inner wall of the liquefied natural gas storage tank in the tank top structure does not have continuous and stable geometric characteristics in space. If this area is directly included in the division process of the tank top plane, it is easy to cause local geometric distortion or region overlap in the plane division result, thus affecting the accuracy of subsequent zoning analysis. By identifying and excluding the non-projection area before plane division, it can be ensured that the tank top plane consists only of effective areas that are geometrically continuous and structurally complete. On this basis, the tank top plane is divided into circumferential sectors along the circumference of the storage tank, so that each circumferential sector corresponds to a regional unit in the tank top structure with clear spatial significance and stable structural characteristics, thus providing a reliable spatial basis for subsequent multi-source data acquisition, edge convergence, and fusion analysis based on circumferential sectors.

[0043] S3. Perform edge-side distributed collaborative aggregation operation on the circumferential sector to obtain the baseline quantity set; In an embodiment of the present invention, an edge-side distributed collaborative aggregation operation is performed on the circumferential sector to obtain a baseline quantity set, including: Deploy edge acquisition terminals in each circumferential sector; Specifically, the process of deploying edge acquisition terminals in each circumferential sector includes determining the spatial range of each circumferential sector on the tank top structure based on the circumferential sector division results of the tank top plane, and setting up edge acquisition terminals for data acquisition within the corresponding spatial range, so that each edge acquisition terminal corresponds one-to-one with the corresponding circumferential sector in spatial position, thereby ensuring that the data collected subsequently can accurately reflect the operating status of the tank top structure within its respective circumferential sector.

[0044] Temperature value sequences for each circumferential sector are collected using an edge acquisition terminal; Specifically, temperature samples are taken from the surface of the tank top structure or its adjacent structural parts at the edge acquisition terminal, and multiple temperature values ​​are continuously acquired at predetermined sampling time intervals. The acquired temperature values ​​are recorded and arranged in chronological order to form a temperature value sequence that characterizes the temperature change of the circumferential sector. This temperature value sequence is used to reflect the thermal state changes of the tank top structure within the corresponding circumferential sector.

[0045] The temperature value sequences of each circumferential sector are locally aggregated, and all temperature value sequences are aggregated to obtain the circumferential average temperature of the tank top structure. Specifically, the edge acquisition terminal corresponding to each circumferential sector first performs outlier removal processing on its continuously acquired temperature value sequence. The temperature value sequence is first sorted by numerical value, and the value in the middle position after sorting is taken as the median of the sequence. Then, the absolute difference between each temperature value and the median is calculated. All absolute differences are sorted by numerical value, and the value in the middle position is taken as the absolute deviation of the median. Three times the absolute deviation of the median is set as the anomaly judgment threshold. If the absolute difference between a temperature value and the median exceeds the threshold, it is judged as an outlier and removed from the sequence. After retaining all valid temperature values, their arithmetic mean is calculated as the temperature representative value of the corresponding circumferential sector to complete the local aggregation processing. Then, the temperature representative values ​​obtained by the local aggregation of all circumferential sectors are collected, the total number of temperature representative values ​​participating in the aggregation operation is counted, all temperature representative values ​​are accumulated to obtain the temperature sum, and the temperature sum is divided by the total number of temperature representative values. The quotient value is the circumferential average temperature of the tank top structure.

[0046] Specifically, the edge acquisition terminal refers to an acquisition device deployed within the corresponding circumferential sector to acquire relevant operational data of that circumferential sector nearby, and its location corresponds spatially to the circumferential sector; the temperature value sequence refers to a set of temperature data arranged in chronological order, acquired by the edge acquisition terminal during continuous sampling, reflecting the temperature changes of the tank top structure within the corresponding circumferential sector; the local aggregation processing refers to the process of organizing and aggregating the temperature value sequences acquired within the same circumferential sector near the acquisition location to form data results that can represent the temperature characteristics of that circumferential sector; the aggregation operation refers to the calculation process of uniformly processing temperature value sequences from different circumferential sectors to comprehensively reflect the overall temperature distribution characteristics of the tank top structure; the circumferential average temperature refers to the numerical result obtained after aggregating the temperature data corresponding to all circumferential sectors, used to characterize the overall temperature level of the liquefied natural gas storage tank top structure in the circumferential direction.

[0047] The strain value sequence of the circumferential sector is acquired through the edge acquisition terminal; Specifically, based on the circumferential sector division of the tank top plane, the spatial range of each circumferential sector on the tank top structure is determined one by one. At least three evenly distributed strain sampling points are selected on the surface of the tank top structure corresponding to each circumferential sector. Strain sensors are fixedly installed at each sampling point. The signal output terminals of all strain sensors are physically connected to the edge acquisition terminal of the corresponding circumferential sector. The sampling time interval is set to once every 10 seconds. The edge acquisition terminal triggers the strain sensors to continuously sample at this interval. Each sampling acquires the strain data at the corresponding position. All the sampled strain data are recorded and arranged in chronological order of acquisition time. The resulting continuous data set is the strain value sequence of the corresponding circumferential sector.

[0048] The strain value sequences of each circumferential sector are locally aggregated, and all strain value sequences are aggregated to obtain the circumferential average strain of the tank top structure. Specifically, the edge acquisition terminal corresponding to each circumferential sector first performs outlier removal processing on its recorded strain value sequence. The strain value sequence is then sorted by numerical value, and the value in the middle position after sorting is selected as the median of the sequence. The absolute difference between each strain data and the median is calculated. All absolute differences are sorted by numerical value, and the value in the middle position is selected as the absolute deviation of the median. Three times the absolute deviation of the median is set as the outlier judgment threshold. If the absolute difference between a strain data and the median exceeds this threshold, it is removed from the sequence. After retaining all valid strain data that meet the requirements, the arithmetic mean of these valid data is calculated. This average value is used as the strain representative value of the corresponding circumferential sector to complete the local aggregation processing. Subsequently, the strain representative values ​​of all circumferential sectors are collected, the total number of strain representative values ​​involved in the calculation is counted, and all strain representative values ​​are summed to obtain the strain data sum. The quotient obtained by dividing the strain data sum by the total number of strain representative values ​​is the circumferential average strain value of the tank top structure.

[0049] The evaporation gas indicator value sequence of each circumferential sector is collected through the edge acquisition terminal; Specifically, based on the circumferential sector division of the tank top plane, the location of the connecting pipe in the gas phase space of the tank top corresponding to each circumferential sector is determined one by one. A fixed sampling point is selected at the connecting pipe of each circumferential sector, and an evaporating gas flow sensor is installed at the sampling point. The signal output terminal of the sensor is connected to the edge acquisition terminal of the corresponding circumferential sector. The sampling time interval is set to once every 15 seconds. The edge acquisition terminal triggers the sensor to continuously sample at this interval. Each sampling acquires the flow data of the evaporating gas in the corresponding pipe. All the sampled flow data are recorded and arranged in chronological order of acquisition time. The resulting continuous data set is the evaporating gas indication value sequence of the corresponding circumferential sector.

[0050] The evaporation gas indicator value sequences of each circumferential sector are locally aggregated, and all evaporation gas indicator value sequences are aggregated to obtain the uniform distribution value of evaporation gas in the tank top structure. Specifically, the edge acquisition terminal corresponding to each annular sector first performs outlier removal processing on the sequence of evaporating gas indicator values ​​it records. The sequence is then sorted by numerical value, and the value in the middle position after sorting is selected as the median of the sequence. The absolute difference between each evaporating gas indicator data and this median is calculated. All absolute differences are sorted by numerical value, and the value in the middle position is selected as the absolute deviation of the median. 2.5 times the absolute deviation of the median is set as the anomaly judgment threshold. If the absolute difference between a data and the median exceeds this threshold, it is removed from the sequence. After retaining all valid data that meet the requirements, the arithmetic mean of these valid data is calculated. This average is used as the representative value of evaporating gas for the corresponding annular sector to complete the local aggregation processing. Subsequently, the representative values ​​of evaporating gas for all annular sectors are collected, the total number of representative values ​​involved in the calculation is counted, and all representative values ​​are summed to obtain the total evaporating gas data. The sum is divided by the total number of representative values, and the quotient is the uniform distribution value of evaporating gas for the tank top structure.

[0051] Specifically, the strain value sequence refers to a set of strain data arranged in chronological order, acquired by edge acquisition terminals arranged in each circumferential sector during continuous sampling, reflecting the deformation of the tank top structure under operating conditions. This data describes the minute deformation changes of the tank top structure at different locations over time. The circumferential average strain value is a numerical result obtained by collecting and uniformly processing the strain data corresponding to all circumferential sectors, used to characterize the overall deformation level of the liquefied natural gas storage tank top structure in the circumferential direction. The evaporation gas indication value sequence refers to a set of continuous data collected by edge acquisition terminals in each circumferential sector, reflecting the generation or change of liquefied natural gas evaporation gas. This data is related to the thermal state and operating conditions of the tank top structure. The evaporation gas uniform distribution value is a numerical result obtained by uniformly processing the evaporation gas indication data corresponding to all circumferential sectors, used to characterize the overall level of evaporation gas distribution in the circumferential direction of the liquefied natural gas storage tank top structure.

[0052] Specifically, the computational tasks are pushed down to the edge, closer to the data source. Edge acquisition terminals are deployed near each circumferential sector of the tank top structure. Each terminal independently collects real-time data such as temperature, strain, evaporation, and gas in its corresponding sector and performs calculations such as outlier removal, data cleaning, and aggregation locally. There is no need to transmit all raw data to a remote centralized processing center. The local processing results are then shared through a collaborative communication mechanism between the edge acquisition terminals. The processed data from all sectors are aggregated and calculated to form a set of benchmark quantities that can characterize the overall operating status of the tank top structure. The entire process, through distributed deployment and collaborative computation on the edge side, achieves localized data processing and efficient fusion, reducing data transmission latency and network bandwidth consumption.

[0053] The circumferential average temperature, circumferential average strain, and uniform distribution of evaporated gas are collected as the reference values ​​for the circumferential sector.

[0054] S4. Calculate the index value of the circumferential sector based on the benchmark set, and select the target sector based on the index value and the sign comparison result. In an embodiment of the present invention, calculating the index value of the circumferential sector based on the reference quantity set includes: The temperature values ​​of the circumferential sectors are calculated from the temperature value sequence of the circumferential sectors; The temperature deviation is obtained by subtracting the average circumferential temperature value from the reference set from the temperature value of the circumferential sector. Specifically, for a single circumferential sector, the statistical period of the circumferential sector is first determined. All data within this statistical period are extracted from the temperature value sequence corresponding to the sector. These data are then checked for normal fluctuation range, and valid data that conforms to the operating specifications for the top temperature of this type of liquefied natural gas storage tank are retained. The arithmetic mean of these valid data is then calculated, and this mean is the temperature value of the circumferential sector. This step transforms the continuous temperature value sequence into a single characteristic value that can represent the current temperature state of the sector. The step of subtracting the circumferential average temperature value from the benchmark set is based on the logical premise that the circumferential average temperature value is the overall benchmark state of the temperature of all circumferential sectors of the tank top structure. By performing difference calculations, the temperature state of a single circumferential sector is compared with the overall temperature benchmark of the tank top. The resulting temperature deviation can directly characterize the degree of deviation of the temperature of the circumferential sector from the overall temperature level of the tank top. The two steps are sequentially connected, first completing the transformation of the sequence data into sector characteristic values, and then obtaining the deviation characteristics by comparing with the overall benchmark value, providing basic data support for subsequent comprehensive evaluation of the operating status of the circumferential sector.

[0055] The strain values ​​of the circumferential sector are calculated from the strain value sequence of the circumferential sector; The strain deviation is obtained by subtracting the strain value of the circumferential sector from the average strain value in the reference set. Specifically, all data within the statistical period are extracted from the strain value sequence corresponding to the sector. These data are then verified to be within the normal fluctuation range of the tank top structure strain. Valid data that conforms to the tank top deformation operation specifications for this type of liquefied natural gas storage tank are retained. The arithmetic mean of these valid data is then calculated, and this mean is the strain value of the circumferential sector. This step transforms the continuously collected strain value sequence into a single characteristic value that can represent the current structural deformation state of the sector. The step of subtracting the strain value from the circumferential average strain value in the benchmark set is based on the logical premise that the circumferential average strain value is the overall benchmark state of the deformation of all circumferential sectors of the tank top structure. By performing difference calculations, the deformation state of a single circumferential sector is compared with the overall deformation benchmark of the tank top. The resulting strain deviation can directly characterize the degree of deviation of the structural deformation of the circumferential sector from the overall deformation level of the tank top. The two steps are sequentially connected. First, the sequence data is transformed into sector deformation characteristic values. Then, the deformation deviation characteristics are obtained by comparing with the overall benchmark value, providing key deformation dimension data support for subsequent comprehensive evaluation of the operating status of the circumferential sector.

[0056] Specifically, the temperature value of the circumferential sector refers to the numerical result calculated based on the temperature value sequence, representing the temperature state of the circumferential sector at the current moment or within the statistical period; the reference quantity set refers to the data set composed of the circumferential average temperature, the circumferential average strain, and the uniform distribution value of evaporated gas, used to characterize the overall operating reference state of the tank top structure in the circumferential direction; the temperature deviation refers to the difference between the temperature value of a certain circumferential sector and the circumferential average temperature, which is used to characterize the degree of deviation of the temperature state of the circumferential sector from the overall reference state; the strain value of the circumferential sector refers to the numerical result calculated based on the strain value sequence, representing the structural deformation state of the circumferential sector at the current moment or within the statistical period; the circumferential average strain refers to the numerical result obtained by uniformly processing the strain data corresponding to all circumferential sectors, used to characterize the overall deformation level of the tank top structure in the circumferential direction; the strain deviation refers to the difference between the strain value of a certain circumferential sector and the circumferential average strain, used to reflect the deviation of the structural deformation state of the circumferential sector from the overall reference state.

[0057] The evaporation gas indicator values ​​of the circumferential sector are calculated from the evaporation gas indicator value sequence of the circumferential sector; The evaporation gas indication deviation is obtained by subtracting the evaporation gas uniform distribution value in the reference quantity set from the evaporation gas indication value of the circumferential sector. In an embodiment of the present invention, the formula for calculating the evaporation gas indication deviation is as follows: ; In the formula, It is an evaporation gas indication deviation. This is the value of the evaporated gas indicator. It is the value of uniform distribution of evaporated gas.

[0058] Specifically, all data within the statistical period are extracted from the evaporation gas indicator value sequence corresponding to the sector. These data are then verified against the normal fluctuation range of evaporation gas flow rate at the top of the liquefied natural gas tank. Valid data conforming to the gas phase space operation specifications for this type of tank are retained. The arithmetic mean of these valid data is then calculated; this mean is the evaporation gas indicator value for the circumferential sector. This step transforms the continuously collected evaporation gas indicator value sequence into a single characteristic value representing the current evaporation gas state of the sector. Subtracting the evaporation gas uniform distribution value from the benchmark set from the evaporation gas indicator value is based on the logical premise that the uniform distribution of evaporation gas is the overall benchmark state of the evaporation gas distribution in all circumferential sectors of the tank top structure. Through difference calculation, the evaporation gas state of a single circumferential sector is compared with the overall evaporation gas distribution benchmark at the tank top. The resulting evaporation gas indicator deviation directly characterizes the degree of deviation of the evaporation gas state of the circumferential sector relative to the overall distribution level at the tank top. First, the conversion from sequence data to sector evaporation gas characteristic values ​​is completed. Then, the evaporation gas deviation characteristics are obtained by comparing with the overall benchmark value, providing data support for the evaporation gas dimension for subsequent comprehensive evaluation of the operating status of the circumferential sector.

[0059] The sign of the evaporating gas indication deviation is extracted to obtain the deviation sign value; The absolute values ​​of temperature deviation, strain deviation, and deviation sign are multiplied sequentially to obtain the index values ​​of the circumferential sector.

[0060] Specifically, the process of extracting the sign of the evaporating gas indication deviation involves determining the positive or negative value of the deviation. If the deviation is greater than 0, the extracted sign value is 1; if the deviation is less than 0, the extracted sign value is -1. This deviation sign value is used to clarify the direction of deviation of the evaporating gas state in the circumferential sector relative to the overall baseline. The step of multiplying the absolute values ​​of the temperature deviation and strain deviation sign values ​​sequentially involves first calculating the absolute value of the temperature deviation after removing its positive or negative attributes, then calculating the absolute value of the strain deviation after removing its positive or negative attributes, multiplying the absolute values ​​of the temperature and strain deviations, and finally multiplying the product by the deviation sign value to obtain the index value for the circumferential sector. By first clarifying the direction of the evaporating gas deviation through sign extraction, and then combining the magnitudes of the temperature and strain deviations, the deviation characteristics of the three dimensions are integrated into a single index value through multiplication. This comprehensively characterizes the degree of deviation of the circumferential sector relative to the overall baseline of the tank top in terms of thermal structural deformation and evaporating gas state.

[0061] Specifically, the evaporation gas indication value of the circumferential sector refers to the numerical result calculated based on the evaporation gas indication value sequence, representing the evaporation gas state of the circumferential sector at the current moment or within a statistical period; the evaporation gas indication deviation refers to the difference between the evaporation gas indication value of a certain circumferential sector and the uniform distribution value of evaporation gas, used to reflect the degree of deviation of the evaporation gas state of the circumferential sector from the overall baseline state; the deviation sign quantity refers to the sign result extracted based on the positive and negative attributes of the evaporation gas indication deviation, used to characterize the direction of deviation of the evaporation gas state of the circumferential sector from the overall baseline state; The absolute value of temperature deviation refers to the magnitude of the temperature deviation of the circumferential sector after removing the positive and negative directions, and is used to characterize the magnitude of the temperature change of the circumferential sector relative to the overall reference state; the absolute value of strain deviation refers to the magnitude of the strain deviation of the circumferential sector after removing the positive and negative directions, and is used to characterize the magnitude of the structural deformation of the circumferential sector relative to the overall reference state; the index value is the result obtained by combining the absolute values ​​of temperature deviation, strain deviation, and the deviation sign value in sequence, and is used to comprehensively reflect the degree of deviation of the corresponding circumferential sector in terms of thermal state, structural deformation state, and evaporating gas state.

[0062] In embodiments of the present invention, the target sector is selected based on the index value and symbol comparison result, including: The average index value is obtained by averaging the index values ​​of all circumferential sectors. The difference between the index value and the average index value of the circumferential sector is calculated to obtain the index difference. Specifically, the process begins by counting the total number of circumferential sectors obtained from dividing the tank top structure. This number represents the total number of index values ​​to be calculated. Then, the index values ​​for each circumferential sector are summed one by one to obtain the total index values ​​for all circumferential sectors. This sum is then divided by the total number of circumferential sectors to obtain the arithmetic mean, which is the average index value. This value is used to characterize the overall baseline level of the comprehensive operating status of all circumferential sectors of the tank top structure. The process of calculating the difference between the index values ​​of the circumferential sectors and the average index value is as follows: for each circumferential sector, the average index value is subtracted from the index value corresponding to that sector. The resulting difference is the index difference value for that circumferential sector. First, the dispersed index values ​​of multiple circumferential sectors are integrated into a baseline value representing the overall comprehensive status of the tank top through averaging calculations, providing a reference standard for the status assessment of individual sectors. Then, the comprehensive status of a single circumferential sector is compared with the overall baseline through difference calculations, quantifying the degree of deviation of that sector from the overall comprehensive status of the tank top. This provides a basic quantitative basis for subsequently selecting target sectors with significant change characteristics.

[0063] Specifically, the average index value refers to the numerical result obtained by uniformly calculating the index values ​​corresponding to all circumferential sectors, which is used to characterize the overall reference state level of the tank top structure in the circumferential direction; the index difference refers to the difference between the index value of a certain circumferential sector and the average index value, which is used to reflect the degree of deviation of the comprehensive state of the circumferential sector from the overall reference state; and it is used to reflect the correspondence between the direction of thermal state change and the direction of structural deformation change of the circumferential sector.

[0064] Sign comparison was performed on temperature deviation and strain deviation to obtain the sign comparison results; Specifically, for a single circumferential sector, first, the temperature deviation value corresponding to that sector is obtained, and its positive or negative attribute is determined. If the temperature deviation value is greater than 0, the sign of the temperature deviation is determined to be positive; if the temperature deviation value is less than 0, the sign of the temperature deviation is determined to be negative. Next, the strain deviation value corresponding to that sector is obtained, and its positive or negative attribute is determined in the same way. If the strain deviation value is greater than 0, the sign is positive; if it is less than 0, the sign is negative. Then, the sign of the temperature deviation of that sector is compared with the sign of the strain deviation. If the two signs are the same, the sign comparison result corresponding to that circumferential sector is determined to be consistent; if the two signs are opposite, the sign comparison result corresponding to that circumferential sector is determined to be opposite.

[0065] Specifically, the symbol comparison result refers to the judgment result obtained by comparing the positive and negative attributes of the temperature deviation and strain deviation corresponding to the same circumferential sector after determining their positive and negative attributes. This judgment result is used to characterize whether the direction of temperature change and the direction of structural strain change in the circumferential sector show a consistent or opposite relationship, thereby reflecting the corresponding characteristics between thermal state change and structural deformation change in the circumferential sector. The symbol comparison result serves as one of the important bases for subsequent screening of target sectors and is used to assist in judging whether the comprehensive change in the circumferential sector has a clear directional characteristic.

[0066] Based on the index difference and sign comparison results, the target sector is selected from all circumferential sectors.

[0067] Specifically, first, historical data on the index differences generated by all circumferential sectors during the continuous 30-day operation of the top structure of this type of liquefied natural gas storage tank are collected. After sorting these historical data, the first and last 10% of extreme data are removed. The arithmetic mean and standard deviation of the remaining data are calculated, and the 2% of the standard deviation is then used as the basis for further analysis. The deviation threshold is set as a multiple, and the arithmetic mean plus the deviation threshold is used as the upper limit critical value of the index difference, while the arithmetic mean minus the deviation threshold is used as the lower limit critical value of the index difference. Then, for each circumferential sector, its corresponding index difference is first obtained, and it is determined whether the index difference is greater than the upper limit critical value or less than the lower limit critical value. If the index difference is between the upper limit critical value and the lower limit critical value, it is determined that the overall deviation of the sector is not significant and is not included in the candidate range. If the index difference exceeds the upper limit critical value or is lower than the lower limit critical value, the sign comparison result corresponding to the sector is further obtained. When the sign comparison result shows that the signs of temperature deviation and strain deviation are consistent, the circumferential sector is marked as a candidate target. After completing the index difference judgment and sign comparison result association operation for all circumferential sectors, all marked candidate targets are summarized. These summarized circumferential sectors are the selected target sectors, which can be used as the key monitoring and analysis areas for the thermal state and deformation state of the tank top structure in the future.

[0068] Specifically, the target sector refers to the circumferential region with significant comprehensive change characteristics selected from all circumferential sectors after dividing the tank top structure into multiple circumferential sectors. This is based on the differences between the corresponding index values ​​of each circumferential sector and the average index value, and after comprehensive judgment combined with the sign comparison results of temperature deviation and strain deviation. This circumferential region shows a significant deviation from the overall baseline state in terms of thermal state, structural deformation state, or the correspondence between the two. It is used as a spatial unit for subsequent fixed-point inspection, key analysis, or operational attention.

[0069] Specifically, the purpose of selecting target sectors is that after dividing the tank top structure into circumferential sections, different circumferential sectors often exhibit differences in thermal state, structural deformation state, and the corresponding relationship between the two. If all circumferential sectors are analyzed or inspected in the same way, it is easy to cause the analysis scope to be too large and lack specificity. By selecting target sectors based on index difference and sign comparison results, areas that show significant deviations from the overall baseline state in terms of comprehensive state can be identified from all circumferential sectors. This allows subsequent fixed-point inspections, key analyses, or operational attention to be focused on circumferential areas with representative and potential risk characteristics, thereby improving the pertinence and effectiveness of tank top structure analysis and inspection.

[0070] S5. Perform point-to-point checks on the target sector based on the data fusion value of the target sector.

[0071] In an embodiment of the present invention, a point-to-point check of the target sector is performed based on the data fusion value of the target sector, including: Data fusion is performed on temperature deviation and strain deviation to obtain fused data values; In an embodiment of the present invention, the formula for calculating the data fusion value is as follows: ; In the formula, It's a temperature deviation. It is strain deviation. , These are the coefficients of a linear combination. It is the data fusion value.

[0072] Specifically, the data fusion value is calculated by linearly combining temperature deviation and strain deviation. This is based on the objective coupling relationship between thermal state change and structural deformation response. Temperature deviation reflects the degree of change in the thermal state of a local area of ​​the tank top structure relative to the overall thermal state, while strain deviation reflects the structural deformation response of the same spatial area under heating conditions. In engineering practice, structural materials often exhibit approximately linear thermally induced deformation characteristics under temperature changes. Therefore, by assigning linear combination coefficients to temperature deviation and strain deviation and performing superposition calculations, the influence of both on the overall state of the tank top structure can be uniformly mapped into a scalar result. This scalar result can simultaneously reflect the combined effect of heat input change and structural response change, thus forming a quantitative description of the overall state of the local circumferential sector. This quantitative result is used as the data fusion value for subsequent analysis and judgment.

[0073] Specifically, historical operating data of the LNG storage tank top structure over six consecutive months was first selected. Temperature deviation and structural condition deviation samples were extracted for each circumferential sector under different operating conditions. Simultaneously, a comprehensive condition assessment of the tank top structure under the corresponding operating conditions was collected. This assessment, obtained through manual fixed-point inspection combined with high-precision structural condition monitoring equipment, reflects the actual operating state under the coupling effect of thermal and structural forces on the tank top. Then, using temperature deviation and structural condition deviation as input information and the actual comprehensive condition assessment results as output information, a linear regression method was used to fit these sample data. The goal of the fitting process was to minimize the sum of squared errors between the input information and the output information. Two weighted parameters were then calculated to minimize this sum of squared errors, with the parameter corresponding to the temperature deviation being... The weighting parameters are one of the linear combination coefficients, and the weighting parameter corresponding to the structural state deviation is the other linear combination coefficient. Finally, select independent operating data of another month for the tank top structure, substitute the temperature deviation and structural state deviation into the calculation formula of the data fusion value, and then compare the calculated result with the actual comprehensive state assessment result of the tank top structure in the corresponding period. If the deviation between the two is within the allowable error range of the state assessment of this type of tank top structure, then the two weighting parameters obtained are determined as the final linear combination coefficients. If the deviation exceeds the allowable range, then select more historical operating data for a longer period of time and repeat the above sample fitting and data verification process until the two weighting parameters obtained can make the deviation between the calculated result and the actual comprehensive state assessment result meet the accuracy requirements of the state assessment of this type of tank top structure.

[0074] The data fusion value is defined as the additional heat index value of the target sector; Specifically, the additional heat index value is used to reflect the degree of additional heat influence that the target sector bears or transmits relative to the overall state of the tank top structure. This index can comprehensively reflect the local heat input anomaly and the resulting structural response characteristics.

[0075] Specifically, the data fusion value is defined as the additional heat index value of the target sector because it comprehensively reflects both the temperature change characteristics of the target sector relative to the overall baseline state and the structural strain response characteristics caused by the temperature change. Both of these characteristics point to the change in the heat input or transfer level in the region relative to the overall state. By giving this comprehensive quantitative result an engineering meaning of additional heat, the dispersed temperature information and structural response information can be unified into an index form that characterizes the strength of local thermal effects. This allows the non-uniform thermal effects experienced by the target sector to be described by a single index, thus providing an intuitive and directional basis for subsequent inspection and analysis of the target sector.

[0076] The target sector is inspected based on the additional heat index value.

[0077] Specifically, the fixed-point inspection refers to the targeted inspection or verification operation of the tank top structure area corresponding to the target sector based on the spatial location indicated by the additional heat index value, in order to confirm whether there is an actual situation in the area corresponding to the change in thermal state or structural state, thereby achieving precise focus on key areas of the tank top structure.

[0078] Specifically, first, the additional heat index values ​​of all target sectors are sorted in descending order to determine the inspection priority of each target sector. Target sectors with higher values ​​are inspected first. Then, inspection tools such as contact temperature detectors, portable thermal imagers, and strain gauges are prepared. Based on the division marks of the circumferential sectors on the tank top, the target sector spatial area corresponding to the additional heat index value is located on the surface of the tank top structure. First, the portable thermal imager is used to scan the entire surface of the target sector to obtain a real-time temperature distribution image of the area. This image is compared with a baseline image of the overall temperature distribution of the tank top structure, and local sub-regions in the image where the temperature is significantly higher than the baseline are marked. Then, the contact temperature detector is used to select at least five evenly distributed measurement points within the marked local sub-regions, and the real-time temperature value of each point is measured and recorded to verify the actual temperature deviation of the area. The degree of thermal impact is determined by whether it matches the thermal impact level corresponding to the additional heat index value. Then, the detection probe of the strain gauge is attached to the structural surface of the target sector to collect real-time strain data of the area. This data is compared with the overall strain reference data of the tank top structure to confirm whether the structural strain response of the area is consistent with the thermally induced deformation characteristics reflected by the additional heat index value. Then, the structural surface condition of the target sector is visually observed to check for physical damage related to heat effects, such as coating cracking, structural deformation, or loose sealing components. Finally, the real-time temperature data, strain data, and visual inspection results are summarized and compared with the thermal impact description corresponding to the additional heat index value to determine whether there is an abnormal heat input or structural damage risk in the target sector. At the same time, all inspection data and observation results are recorded to form a target sector fixed-point inspection report, which serves as the basis for subsequent tank top structure operation adjustment or maintenance work.

[0079] The above description is only a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and inventive concept of the present invention, should be covered within the scope of protection of the present invention.

Claims

1. A multi-source data fusion method for receiving stations based on edge computing, characterized in that, Includes the following steps: S1. Generate the circumferential position of the annular gap, and construct a non-projected region based on the three-dimensional spatial coordinate data and circumferential position of the annular gap; S2. Divide the tank top plane of the tank top structure into multiple circumferential sectors based on the non-projection area; S3. Perform edge-side distributed collaborative aggregation operation on the circumferential sector to obtain the baseline quantity set; S4. Calculate the index value of the circumferential sector based on the benchmark set, and select the target sector based on the index value and the sign comparison result. S5. Perform point-to-point checks on the target sector based on the data fusion value of the target sector.

2. The multi-source data fusion method for receiving stations based on edge computing according to claim 1, characterized in that, The circumferential positions for generating the annular gap include: The top structure of the liquefied natural gas storage tank in the positioning receiving station; The annular gap between the suspension plate in the top structure of the positioning tank and the inner tank wall of the liquefied natural gas storage tank; Obtain the three-dimensional spatial coordinate data of the annular gap; Angle calculations are performed on the three-dimensional spatial coordinate data of the annular gap to obtain the circumferential position of the annular gap.

3. The multi-source data fusion method for receiving stations based on edge computing according to claim 2, characterized in that, A non-projected region is constructed based on the three-dimensional spatial coordinate data and circumferential position of the annular gap, including: The radial distance of the annular gap is calculated by analyzing its three-dimensional spatial coordinate data to obtain its radial position. Obtain the three-dimensional geometric data of the tank top structure; Based on the circumferential and radial positions, each three-dimensional geometric point in the three-dimensional geometric data is determined and marked to obtain the gap point; The gap points are aggregated to form a non-projected region.

4. The multi-source data fusion method for receiving stations based on edge computing according to claim 2, characterized in that, Based on the non-projection area, the tank top plane of the tank top structure is divided into multiple circumferential sectors, including: Geometric constraint processing is performed on the three-dimensional geometric data of the tank top structure based on the non-projection region to determine the reference projection range; Within the reference projection range, projection calculations are performed on the three-dimensional geometric data of the tank top structure to obtain two-dimensional projection data; The two-dimensional projection data is reconstructed into a plane to obtain the tank top plane. The top plane of the tank is divided into multiple circumferential sectors.

5. The multi-source data fusion method for receiving stations based on edge computing according to claim 1, characterized in that, Perform edge-side distributed collaborative aggregation operations on the circumferential sectors to obtain a set of baseline quantities, including: Deploy edge acquisition terminals in each circumferential sector; Temperature value sequences for each circumferential sector are collected using an edge acquisition terminal; The temperature value sequences of each circumferential sector are locally aggregated, and all temperature value sequences are aggregated to obtain the circumferential average temperature of the tank top structure. The strain value sequence of the circumferential sector is acquired through the edge acquisition terminal; The strain value sequences of each circumferential sector are locally aggregated, and all strain value sequences are aggregated to obtain the circumferential average strain of the tank top structure. The evaporation gas indicator value sequence of each circumferential sector is collected through the edge acquisition terminal; The evaporation gas indicator value sequences of each circumferential sector are locally aggregated, and all evaporation gas indicator value sequences are aggregated to obtain the uniform distribution value of evaporation gas in the tank top structure. The circumferential average temperature, circumferential average strain, and uniform distribution of evaporated gas are collected as the reference values ​​for the circumferential sector.

6. The multi-source data fusion method for receiving stations based on edge computing according to claim 5, characterized in that, The index values ​​of the circumferential sector are calculated based on the baseline set, including: The temperature values ​​of the circumferential sectors are calculated from the temperature value sequence of the circumferential sectors; The temperature deviation is obtained by subtracting the average circumferential temperature value from the reference set from the temperature value of the circumferential sector. The strain values ​​of the circumferential sector are calculated from the strain value sequence of the circumferential sector; The strain deviation is obtained by subtracting the strain value of the circumferential sector from the average strain value in the reference set. The evaporation gas indicator values ​​of the circumferential sector are calculated from the evaporation gas indicator value sequence of the circumferential sector; The evaporation gas indication deviation is obtained by subtracting the evaporation gas uniform distribution value in the reference quantity set from the evaporation gas indication value of the circumferential sector. The sign of the evaporating gas indication deviation is extracted to obtain the deviation sign value; The absolute values ​​of temperature deviation, strain deviation, and deviation sign are multiplied sequentially to obtain the index values ​​of the circumferential sector.

7. The multi-source data fusion method for receiving stations based on edge computing according to claim 6, characterized in that, The formula for calculating the deviation of the evaporating gas indication is as follows: ； In the formula, It is an evaporation gas indication deviation. This is the value of the evaporated gas indicator. It is the value of uniform distribution of evaporated gas.

8. The multi-source data fusion method for receiving stations based on edge computing according to claim 6, characterized in that, Target sectors were selected based on the index values ​​and sign comparison results, including: The average index value is obtained by averaging the index values ​​of all circumferential sectors. The difference between the index value and the average index value of the circumferential sector is calculated to obtain the index difference. Sign comparison was performed on temperature deviation and strain deviation to obtain the sign comparison results; Based on the index difference and sign comparison results, the target sector is selected from all circumferential sectors.

9. The multi-source data fusion method for receiving stations based on edge computing according to claim 1, characterized in that, Target sector point inspection is performed based on the data fusion values ​​of the target sector, including: Data fusion is performed on temperature deviation and strain deviation to obtain fused data values; The data fusion value is defined as the additional heat index value of the target sector; The target sector is inspected based on the additional heat index value.

10. The multi-source data fusion method for receiving stations based on edge computing according to claim 9, characterized in that, The formula for calculating the data fusion value is as follows: ； In the formula, It's a temperature deviation. It is strain deviation. , These are the coefficients of a linear combination. It is the data fusion value.

Images