Method and system for calibration of underground tank volume table at a gas station

Abstract

The application provides a method and system for correcting the volume table of a ground tank at a gas station, and relates to the technical field of oil product metering. The application obtains the refueling details, real-time liquid level data and oil unloading data of the ground tank at the gas station within a preset time period. A plurality of sample points including transaction volume and real liquid level change height are extracted from the refueling details and real-time liquid level data through self-matching to determine a volume compensation amount. Based on the sample points and the volume compensation amount, a plurality of intrinsic mode components are obtained. Low-frequency intrinsic mode components are extracted from the intrinsic mode components through energy centroid and superimposed and reconstructed to obtain a target intrinsic sequence. A residual function is constructed based on a pre-constructed ground tank model. The initial geometric parameters of the ground tank model are iteratively optimized using a fractional order gradient optimization differential operator to generate a corrected volume table. The application realizes accurate identification of the geometric parameters of the ground tank and improves the accuracy, reliability and automation level of the correction of the full-range volume table.

Description

Technical Field

[0001] This application relates to the field of oil metering technology, and in particular to a method and system for calibrating the volume meter of the underground tank at a gas station. Background Technology

[0002] With the rapid development of smart refined oil retail, the accuracy of gas station tank volume meters has become crucial for improving inventory control precision and reducing oil product losses. Precise volume calibration technology is not only a prerequisite for achieving automated profit and loss balance, but also an important foundation for ensuring refined asset management and fair trade settlement at gas stations.

[0003] Existing methods primarily involve real-time matching of automatic level gauge data and oil output data, combined with mathematical models such as least squares method or conventional polynomial fitting to correct the tank volume curve. Although some systems have introduced basic data cleaning processes, they still largely rely on static geometric modeling or simple data association logic to derive the mapping relationship between level and volume.

[0004] However, due to complex interferences such as liquid level fluctuations, sensor noise, and ambient temperature differences at gas station sites, existing methods struggle to effectively separate the true volumetric characteristics from dynamic data, leading to significant systematic biases in the fitted models. Furthermore, traditional optimization algorithms struggle to maintain accuracy across the entire volumetric range when dealing with complex nonlinear geometric deviations of gas tanks. Therefore, existing technologies suffer from low accuracy and poor robustness in correcting gas tank volumetric tables under complex dynamic environments. Summary of the Invention

[0005] The purpose of this application is to provide a method and system for calibrating the volumetric gauge of gas station tanks, in order to solve the technical problems of low calibration accuracy and poor robustness of the volumetric gauge in complex dynamic environments in the prior art.

[0006] Firstly, this application provides a method for correcting the volume table of underground gas station tanks, including: Obtain refueling details, real-time liquid level data, and unloading data of the gas station's underground tank within a preset time period. The unloading data includes the unloading volume and unloading height. Multiple sample points, including transaction volume and actual liquid level change height, are extracted from refueling details and real-time liquid level data through self-matching, and the volume compensation amount is determined based on unloading data, refueling details during unloading, and volume table. Based on sample points and volume compensation, a time-frequency signal is constructed using the mapping operator in the multidimensional signal fusion analysis algorithm. The time-frequency signal is then non-recursively decomposed using variational mode decomposition to obtain multiple intrinsic mode components. The low-frequency intrinsic mode components are extracted from the intrinsic mode components by the energy centroid and then superimposed and reconstructed to obtain the target intrinsic sequence representing the relationship between the height and volume of the earth tank. The theoretical volume sequence is determined based on the pre-constructed earth tank model, and the residual function is constructed based on the difference between the target intrinsic sequence and the theoretical volume sequence. The initial geometric parameters of the earth tank model are iteratively optimized using a fractional gradient optimization differential operator to determine the target geometric parameters that minimize the residual function. The target geometric parameters are then mapped to the earth tank model, and a correction volume table is generated through piecewise numerical integration across the entire range.

[0007] Optionally, the refueling details include the transaction time and volume, and the real-time liquid level data includes the real-time sampling height and timestamp. Multiple sample points, including the transaction volume and the actual liquid level change height, are extracted from the refueling details and real-time liquid level data through self-matching. Based on the transaction time, determine multiple transaction time periods corresponding to each transaction volume; Based on the timestamp corresponding to the transaction time period, extract the start sampling height and end sampling height corresponding to each transaction time period from the real-time sampling height; Subtract the corresponding starting sampling height from each termination sampling height to obtain the actual liquid level change height for each trading period. Using the trading time period as an index, each actual liquid level change height is associated with the corresponding trading volume to obtain multiple sample points.

[0008] Optionally, the volume compensation amount is determined based on the unloading data, the refueling details during the unloading period, and the volume table, including: The total sales volume is obtained by summing the volumes of all transactions during the unloading period in the fuel refueling details. Subtract the total sales volume from the amount of oil unloaded on the voucher to obtain the actual oil volume; Using the volume table, look up the first volume corresponding to the oil unloading start height and the second volume corresponding to the oil unloading end height, and subtract the first volume from the second volume to obtain the theoretical oil inlet volume. The volume compensation is obtained by subtracting the theoretical oil intake volume from the actual oil volume.

[0009] Optionally, based on sample points and volume compensation, a time-frequency signal is constructed using the mapping operator in a multidimensional signal fusion analysis algorithm. The time-frequency signal is then non-recursively decomposed using variational mode decomposition to obtain multiple intrinsic mode components, including: A liquid level coordinate axis is constructed based on the real-time sampling height and unloading height. The mapping operator is used to map the transaction volume and volume compensation amount corresponding to each sample point to the liquid level coordinate axis, resulting in multiple fusion points. Based on the height value of each fusion point, all fusion points are sorted in ascending order to obtain an ordered point set. The volume values ​​in the ordered point set are interpolated with equal step size to obtain the time-frequency signal. The spectral distribution of the time-frequency signal is obtained by spectral analysis. The spectral distribution includes multiple sub-signals and the center frequency corresponding to each sub-signal. Each sub-signal includes amplitude information and frequency value. Based on the degree of diffusion of each sub-signal relative to the corresponding center frequency, the energy bandwidth of each sub-signal is calculated, and each center frequency is iteratively adjusted until the sum of the energy bandwidths of all sub-signals meets the preset condition, thus obtaining the updated center frequency. Based on the updated center frequency, multiple optimized sub-components are extracted from the time-frequency signal. The time-frequency signal is extracted by optimizing the sub-components to obtain multiple target sub-signals, which are then used as intrinsic mode components.

[0010] Optionally, low-frequency intrinsic mode components are extracted from the intrinsic mode components through the energy centroid and then superimposed and reconstructed to obtain the target intrinsic sequence representing the relationship between the height and volume of the earth tank, including: Based on the amplitude information in each intrinsic mode component, the frequency values ​​in each intrinsic mode component are weighted and summed to obtain the energy centroid frequency of each intrinsic mode component; The intrinsic mode components corresponding to the energy centroid frequency that is less than a preset frequency threshold are determined as low-frequency intrinsic mode components. Using the height nodes on the liquid level coordinate axis as the horizontal coordinate index, the volume values ​​corresponding to the low-frequency intrinsic mode components are extracted at each height node and accumulated to obtain the composite volume at each height node. The composite volume is then bound to the corresponding height node to obtain the target intrinsic sequence.

[0011] Optionally, a theoretical volume sequence is determined based on a pre-built earthen tank model, and a residual function is constructed based on the difference between the target intrinsic sequence and the theoretical volume sequence, including: The theoretical volume sequence was obtained by performing volume inversion calculations at each height node using the ground tank model. The numerical distance between each theoretical volume in the theoretical volume sequence and the corresponding synthetic volume in the target intrinsic sequence is calculated to obtain the residual set; The sum of squares of all numerical intervals in the residual set is calculated to obtain the cumulative sum. Based on the mapping relationship between the cumulative sum and the initial geometric parameters, the residual function is constructed.

[0012] Optionally, the initial geometric parameters of the earthen tank model are iteratively optimized using a fractional gradient optimization differential operator to determine the target geometric parameters that minimize the residual function. These target geometric parameters are then mapped to the earthen tank model, and a correction volume table is generated through piecewise numerical integration across the entire range, including: The fractional derivatives of the initial geometric parameters are calculated using differential operators to determine the direction and step size of parameter adjustment. Based on the parameter adjustment direction and parameter adjustment step size, the initial geometric parameters are updated to obtain the updated initial geometric parameters. The updated initial geometric parameters are then substituted into the residual function to obtain the residual value. If the residual value is greater than the preset convergence threshold, the fractional derivative of the initial geometric parameters is calculated using the differential operator until the residual value is less than or equal to the preset convergence threshold. The updated initial geometric parameters are then used as the target geometric parameters. Based on the total range of the liquid level coordinate axis, the tank model is divided into height-dimensional micro-elements using the target geometric parameters to obtain multiple volume micro-element values. The volume element values ​​are accumulated and summed segment by segment according to the ascending height order to determine the cumulative volume value corresponding to each height node. The cumulative volume value is then mapped to the corresponding height node to obtain the corrected volume table.

[0013] Secondly, this application provides a calibration system for gas station underground tank volume tables, including: The acquisition module is used to acquire refueling details, real-time liquid level data and unloading data of the gas station's underground tank within a preset time period. The unloading data includes the unloading volume and unloading height. The determination module is used to extract multiple sample points, including transaction volume and actual liquid level change height, from refueling details and real-time liquid level data through self-matching, and determine the volume compensation amount based on unloading data, refueling details during unloading, and volume table. The module is used to construct time-frequency signals based on sample points and volume compensation, using the mapping operator in the multidimensional signal fusion analysis algorithm, and then using variational mode decomposition to perform non-recursive decomposition on the time-frequency signals to obtain multiple intrinsic mode components. The building module is also used to extract low-frequency intrinsic mode components from the intrinsic mode components through the energy centroid and to superimpose and reconstruct them to obtain the target intrinsic sequence representing the relationship between the height and volume of the earth tank; The construction module is also used to determine the theoretical volume sequence based on the pre-built earth tank model, and to construct the residual function based on the difference between the target intrinsic sequence and the theoretical volume sequence; The determination module is also used to iteratively optimize the initial geometric parameters of the earth tank model using a fractional gradient optimization differential operator to determine the target geometric parameters that minimize the residual function, and then map the target geometric parameters to the earth tank model to generate a correction volume table through full-range piecewise numerical integration.

[0014] Thirdly, this application provides an electronic device, comprising: Memory, used to store computer programs; A processor is used to execute a computer program to implement the steps of the method for correcting the gas station tank volume table as described in the first aspect above.

[0015] Fourthly, this application provides a computer-readable storage medium storing a computer program that, when executed by a processor, can implement the steps of the method for correcting the volume table of the gas station tank as described in the first aspect above.

[0016] The calibration method for gas station tank volume tables provided in this application ensures the comprehensiveness of the analysis sample by acquiring refueling details, real-time liquid level data, and unloading data of the gas station tanks within a preset time period. It achieves a high degree of alignment between sales data and liquid level changes, eliminating interference fluctuations during non-trading periods and improving the consistency and reliability of the original data. It overcomes the limitation of traditional optimization algorithms that easily get trapped in local optima, achieving global and accurate capture of complex nonlinear deviations in the tanks, and ensures that the final generated volume table has extremely high calibration accuracy across the entire range through piecewise numerical integration.

[0017] Furthermore, a liquid level coordinate axis is constructed based on the real-time sampling height and unloading height. A mapping operator is used to fuse the sample point volume and volume compensation amount to the coordinate axis, resulting in a fused point set. Then, the point set is transformed into a continuous height-domain time-frequency signal through ascending order arrangement and equal-step interpolation. Subsequently, spectral analysis is performed on the signal, and the energy bandwidth is calculated based on the diffusion degree of the sub-signal center frequency. The center frequency is iteratively adjusted until the bandwidth minimization constraint is met, thereby accurately extracting the optimized sub-components. Finally, frequency stripping is performed using the optimized sub-components to obtain the eigenmode components that characterize the essential geometric properties of the tank. This method solves the information loss caused by the reliance on single data association logic in traditional methods and reduces systematic biases caused by on-site environmental factors. Attached Figure Description

[0018] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0019] Figure 1 A flowchart illustrating the method for correcting the volume table of underground gas station tanks provided in this application embodiment; Figure 2 A flowchart illustrating the method for obtaining intrinsic modal components provided in this application embodiment; Figure 3 A schematic flowchart illustrating the method for generating a correction volume table provided in an embodiment of this application; Figure 4 A schematic diagram of the structure of the calibration system for the gas station underground tank volume table provided in the embodiments of this application; Figure 5This is a schematic diagram of the hardware structure of the electronic device provided in the embodiments of this application. Detailed Implementation

[0020] To address the challenges of separating true volume characteristics from dynamic data in smart gas stations due to liquid level fluctuations, sensor noise, and environmental interference, and the technical bottleneck of traditional fitting optimization algorithms that easily get stuck in local optima and struggle to maintain full-range accuracy when dealing with complex nonlinear geometric deviations of tanks, this application establishes a multi-source feature fusion foundation through self-matching of full-dimensional data and unloading compensation. It then abandons the traditional static polynomial fitting logic and introduces variational mode decomposition (VMD) technology with non-recursive decomposition capabilities to perform multimodal stripping of the constructed signal. Combined with the energy centroid method, it accurately identifies stable low-frequency components characterizing the essential geometric properties of the tank, thereby overcoming systematic deviations caused by environmental noise at their root. Furthermore, to address the nonlinear optimization dilemma caused by tank deformation, a fractional gradient optimization algorithm with global memory and search advantages is used to perform high-dimensional iterative optimization of the tank's geometric parameters to minimize the residual between the measured and theoretical sequences. Finally, through piecewise numerical integration across the entire range, highly robust, full-range-coverage automated calibration of the volumetric table is achieved, solving the problems of low calibration accuracy and poor robustness in complex dynamic environments encountered by existing technologies.

[0021] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0022] The core of this application is to provide a method for correcting the volume table of gas station underground tanks, and a flowchart of one specific implementation is shown below. Figure 1 As shown, the method includes: Step 101: Obtain the refueling details, real-time liquid level data, and unloading data of the gas station's underground tank within a preset time period. The unloading data includes the unloading volume and unloading height.

[0023] In this step, the gas station underground tank refers to a horizontal metal or composite material container buried underground at the gas station for storing petroleum products. The preset time period refers to a specific time interval for data backtracking analysis, which can be a week, a month, or any selected complete operating cycle. Refueling details refer to the transaction records generated when the fuel dispenser sells petroleum products, which can include the end time of each transaction and the volume of petroleum products sold. Real-time liquid level data refers to the liquid level height information inside the underground tank collected by a high-precision liquid level sensor, which can include the liquid level height value and the corresponding sampling time point. Unloading data refers to the metering information generated during the petroleum replenishment process. Voucher unloading volume refers to the theoretical warehousing volume marked on the oil depot delivery document. Unloading height refers to the liquid level height value inside the underground tank when it is in a stable state before and after the unloading operation is triggered.

[0024] In this embodiment, refueling details are first obtained through a communication interface connected to the fuel dispenser, recording each transaction data point of fuel sold, resulting in a set consisting of the transaction time and volume. Simultaneously, a level gauge is used to collect real-time level data at a preset frequency, obtaining a continuous record of the liquid level fluctuations within the tank over time, resulting in a set of data based on the real-time sampling height. and timestamp The sequence is then formed. Next, the unloading data within the preset time period is obtained, and the unloading volume of each delivery task is recorded.

[0025] For example, when performing calibration work on tank number 1 at gas station A, a predetermined time period is set. This refers to the past 10 consecutive days. During this period, the refueling details obtained include multiple records, such as time... Sold Volume and time Sold Volume The real-time liquid level data acquired simultaneously includes height. and its timestamp When the tank undergoes unloading and refueling, the documented unloading volume is: The corresponding unloading height data includes the static initial height before the unloading operation. and static termination height after unloading .

[0026] Step 102: Extract multiple sample points, including transaction volume and actual liquid level change height, from refueling details and real-time liquid level data through self-matching, and determine the volume compensation amount based on unloading data, refueling details during unloading, and volume table.

[0027] In this step, self-matching refers to the processing logic of aligning sales flow data with liquid level fluctuation data over time using a time index. A sample point is a pair of coordinate data representing the relationship between liquid level height change and corresponding volume during a specific transaction. The actual liquid level change height is the difference between the sampled height at the completion of the refueling operation and the sampled height at the start of the operation. The volume table refers to the baseline data pre-stored in the storage device for querying the correspondence between height and volume. The volume compensation amount is the correction value between the actual volume of oil entering the warehouse and the theoretically calculated volume during the unloading process, caused by measurement environmental interference or sensor deviation.

[0028] Step 201: Based on the transaction time, determine multiple transaction time periods corresponding to each transaction volume.

[0029] In this step, the transaction time period refers to the time interval from when the nozzle is lifted and the flow valve is opened until the nozzle is hung up and the flow metering ends.

[0030] In this embodiment of the application, the transaction time corresponding to the settlement of each transaction is first extracted from the refueling details. Next, by retrieving the state change records generated by the refueling pump during operation, the start time corresponding to this transaction can be traced. The final transaction volume The transaction time period is then obtained by matching this. A positive integer, representing the index number of the transaction record, with a value range of 1. , This represents the total number of transaction records within a preset time period.

[0031] For example, for a set of sales records from gas station A, the extracted transaction time period can be represented as a time vector. ,in, These represent the start and end times of the complete transaction cycle, respectively.

[0032] Step 202: Retrieve the corresponding timestamp based on the transaction time period, and extract the start sampling height and end sampling height corresponding to each transaction time period from the real-time sampling height.

[0033] In this step, the initial sampling height refers to the initial oil level position in the tank collected by the level gauge at the start of the refueling operation. The final sampling height refers to the remaining oil level position in the tank collected by the level gauge at the end of the refueling operation.

[0034] In this embodiment, a defined transaction time period is first used as the retrieval index. Then, the time point coinciding with the start and end points of that time period is searched within the real-time sampled height sequence. Considering the asynchrony between the level gauge sampling frequency and the refueling transaction triggering time, this application uses linear interpolation to reconstruct the sampled height: if... Located at adjacent sampling points and Between, the starting sampling height according to and Proportional interpolation calculations were performed to ensure the physical accuracy of the extracted liquid level change height. (Timestamp) When a match is successful, the corresponding height value is extracted. Specifically, in the height sequence... If found equal and equal Then extract the starting sampling height. for and the termination sampling height for ,in, These represent the 1st, 2nd, and 3rd heights in the real-time sampling height sequence, respectively. The and the first A data pair consisting of a timestamp and the corresponding real-time sampling height. This represents a real-time sampled height sequence; A positive integer, representing the total number of sampling points for real-time liquid level data within a preset time period; A positive integer, representing the index number of the sampling point, with a value range of 1. .

[0035] Step 203: Subtract the corresponding starting sampling height from each termination sampling height to obtain the actual liquid level change height for each trading period.

[0036] In this embodiment, the extracted height information is processed using numerical subtraction. The final sampling height within the same transaction time period is then calculated. With the starting sampling height After subtracting, take the absolute value to obtain This indicates the actual drop in liquid level inside the tank during a single refueling transaction. For example, for the first... This transaction, via height vector The difference result obtained is ,in, These represent the starting sampling height and the ending sampling height in the height vector, respectively.

[0037] Step 204: Using the transaction time period as an index, associate each actual liquid level change height with the corresponding transaction volume to obtain multiple sample points.

[0038] In this embodiment, the transaction time period is first used as a logical link. Then, the calculated actual liquid level change height is... With transaction volume Coordinate binding is performed. The resulting data pairs constitute a set of sample points representing the dynamic characteristics of the tank. For example, multiple generated sample points can be represented as a coordinate sequence. , These represent the 1st, 2nd, and 3rd coordinates in the coordinate sequence, respectively. The and the first A data pair consisting of the actual liquid level change height and the corresponding transaction volume. It is a positive integer, representing the total number of transaction records within the preset time period.

[0039] Step 211: Calculate the cumulative value of all transaction volumes during the unloading period in the refueling details to obtain the total sales volume.

[0040] In this step, the unloading period refers to the time span determined by the start and end times of the unloading operation. Total sales volume refers to the total volume of oil sold through the refueling nozzles during the aforementioned period.

[0041] In this embodiment of the application, the start time of oil unloading is first determined by the oil unloading data. and termination time Then, the refueling details set is traversed to filter out all business records whose transaction times fall within that time range. Finally, the volumes of the filtered transactions are accumulated. For example, in a certain oil unloading operation, the collected transaction volume sequence is... ,in, These represent the 1st, 2nd, and 3rd collected data, respectively. The and the first The total sales volume is obtained by summing the transaction volumes of each individual sale. ,in, This indicates the total number of transactions that occurred during the unloading period. A positive integer representing the index number of the transaction record. , Indicates the first during the unloading period The volume of a single transaction sold.

[0042] Step 212: Subtract the total sales volume from the amount of oil unloaded from the voucher to obtain the actual oil volume.

[0043] In this embodiment, the principle of physical conservation is used to calculate the amount of oil entering and leaving the plant. The amount of oil unloaded is recorded on the delivery slip. Subtract the total sales volume lost during the synchronization time Through the formula The actual volume of the oil is calculated. For example, the result can be obtained by processing the voucher value and the cumulative sales value through scalar subtraction. .

[0044] Step 213: Use the volume table to look up the first volume corresponding to the oil unloading start height and the second volume corresponding to the oil unloading end height, and subtract the first volume from the second volume to obtain the theoretical oil inlet volume.

[0045] In this step, the first volume refers to the static oil inventory obtained by querying the original volume table to be corrected based on the liquid level before unloading. The second volume refers to the static oil inventory obtained by querying the original volume table to be corrected based on the liquid level after unloading. The theoretical oil inlet volume refers to the oil inlet volume derived from changes in height under ideal conditions.

[0046] In this embodiment of the application, the starting height is first extracted from the unloading height. and termination height Next, retrieve the uncalibrated volume table or the reference value calculated based on design parameters. Find it through the numerical index. The corresponding first volume as well as The corresponding second volume Using formulas The theoretical oil inlet volume is calculated. Specifically, this can be done using the volume mapping matrix. The volume components are extracted based on the height component index and the difference is calculated to obtain the volume components. .

[0047] Step 214: Subtract the theoretical oil intake volume from the actual oil volume to obtain the volume compensation amount.

[0048] In this embodiment, the error level is determined by comparing the difference between the measured increment and the theoretical increment. The calculated actual oil volume is then used as the basis for this error determination. Subtract the theoretical oil inlet volume Through the formula The final residual value This is the volume compensation amount.

[0049] Step 103: Based on the sample points and volume compensation, construct the time-frequency signal using the mapping operator in the multidimensional signal fusion analysis algorithm, and perform non-recursive decomposition on the time-frequency signal using variational mode decomposition to obtain multiple intrinsic mode components.

[0050] In this step, the mapping operator refers to a mathematical mapping relationship that converts discrete physical sampled values ​​into a set of spatial coordinate points. The time-frequency signal refers to a spatial domain sequence signal with liquid level height as the independent variable and including volume fluctuations. Variational mode decomposition is a processing algorithm based on a non-recursive variational framework that decomposes a complex signal into several components with specific center frequencies. Intrinsic mode components (EMCs) refer to oscillatory functions that reflect the local characteristics of a signal and satisfy specific bandwidth limitations.

[0051] like Figure 2 As shown, Figure 2 This is a flowchart illustrating the method for obtaining intrinsic modal components provided in an embodiment of this application.

[0052] Step 301: Construct a liquid level coordinate axis based on the real-time sampling height and unloading height, and use the mapping operator to map the transaction volume and volume compensation amount corresponding to each sample point to the liquid level coordinate axis to obtain multiple fusion points.

[0053] In this step, the liquid level coordinate axis refers to a one-dimensional spatial reference dimension constructed based on the internal height of the tank. The mapping operator refers to a mathematical transformation rule that achieves data spatial transformation and logical alignment. The fusion point refers to the associated point with height and volume characteristics generated after projecting multi-source discrete business data onto the liquid level coordinate axis.

[0054] In this embodiment of the application, firstly based on the real-time sampled height sequence The extreme value range of the unloading height sequence determines a one-dimensional coordinate axis with height as the only spatial attribute. Then, the mapping operator is used. Sample point sequence Projected onto the coordinate axes. Simultaneously, the volume compensation amount is... This is mapped to the corresponding unloading height range. (Due to the one-dimensional coordinate axis...) Since only a location benchmark is provided, the transaction volume and volume compensation amount need to be bound as feature attributes at that location, so that each height coordinate carries the corresponding volume change information. For example, in the scenario of gas station A, by establishing a mapping relationship between height coordinates and volume increments, a fused point vector composed of multiple discrete height locations and their corresponding volume values ​​is obtained. ,in, These represent the 1st, 2nd, and 3rd points mapped onto the liquid level coordinate axis, respectively. Each fusion point has a specific value determined by directly projecting and binding the actual liquid level change height and the corresponding transaction volume, or the unloading height and the corresponding volume compensation amount in the sample point, onto a specific position on the liquid level coordinate axis X. is a positive integer, representing the total number of fusion points mapped onto the liquid level coordinate axis.

[0055] Step 302: Arrange all fusion points in ascending order based on the height value of each fusion point to obtain an ordered point set. Perform equal-step interpolation calculation on the volume values ​​in the ordered point set to obtain the time-frequency signal.

[0056] In this step, the ordered point set refers to the set of data points arranged in ascending order of liquid level height. Equal-step interpolation calculation refers to an algorithm that fills the estimated values ​​between discontinuous sampling points according to fixed liquid level intervals. The time-frequency signal refers to a continuous volume fluctuation sequence generated by interpolation, with liquid level height as the independent variable.

[0057] In this embodiment of the application, the fusion point vector is first extracted. The height component of each element is determined. All fusion points are sorted using a quicksort algorithm to obtain an ordered set of points. Then, a fixed liquid level increment is set as the interpolation step size. Perform interpolation filling on the volume values ​​of the ordered point set.

[0058] For example, for ordered point set vectors Between adjacent elevation points, by step size The estimated volume is filled in, ultimately resulting in a length that represents the volume as a continuous change in height. Time-frequency signal sequence ,in, They represent the 1st, 2nd, and 3rd points in the ordered set, respectively. The ordered points are the result of extracting the height components of all fused points in the aforementioned fused point vector Q and rearranging them in ascending order using a quicksort algorithm. These represent the first, second, and third times generated by equal-step interpolation of the time-frequency signal sequence, respectively. A continuous volume fluctuation value. Represents a time-frequency signal sequence. is a positive integer representing the total length of the time-frequency signal sequence after interpolation.

[0059] Step 303: Perform spectral analysis on the time-frequency signal to obtain the spectral distribution. The spectral distribution includes multiple sub-signals and the center frequency corresponding to each sub-signal. Each sub-signal includes amplitude information and frequency value.

[0060] In this step, the spectral distribution refers to the set of energy intensities of a signal at different frequency components. A sub-signal refers to a single frequency or narrowband component that makes up the original complex signal. The center frequency is the frequency value at which the sub-signal's energy is most concentrated. Amplitude information refers to the intensity of the sub-signal at a specific frequency. The frequency value refers to the oscillation frequency corresponding to the sub-signal.

[0061] In this embodiment of the application, the time-frequency signal sequence is first analyzed. Performing a Fast Fourier Transform yields a complex sequence that reflects the frequency characteristics of the original signal. Next, for complex sequences... The modulo operation is performed on each complex component, that is, the square root of the sum of the squares of the real and imaginary parts is calculated to obtain the amplitude intensity at each frequency node. The amplitude intensity variation with frequency is used to construct a spectral distribution reflecting the variation of the tank's cross-sectional area with height. The low-frequency amplitude intensity represents the continuous evolution trend of the tank's cross-sectional area, while the high-frequency amplitude intensity represents random interference caused by liquid surface fluctuations or sensor errors. Multiple sub-signals with independent energy peaks are identified from this distribution.

[0062] For example, determining the first peak value through spectral peak retrieval The center frequency of each sub-signal and the corresponding amplitude information The final result is a set of spectral features composed of multiple sub-signal parameters. It is a positive integer, used as a dedicated index number for the frequency domain sub-signals.

[0063] Step 304: Calculate the energy bandwidth of each sub-signal based on the degree of diffusion of each sub-signal relative to the corresponding center frequency, iteratively adjust each center frequency until the sum of the energy bandwidths of all sub-signals meets the preset condition, obtain the updated center frequency, and extract multiple optimized sub-components from the time-frequency signal based on the updated center frequency.

[0064] In this step, energy bandwidth refers to the width of the energy coverage range of the sub-signal in the frequency domain. Preset conditions refer to the criteria for determining whether the iterative process has reached convergence accuracy. Optimized sub-components refer to the high-quality narrowband components determined after variational iterative balancing.

[0065] In this embodiment, the energy bandwidth is first calculated based on the energy distribution of each sub-signal. A penalty factor is introduced to construct a variational constraint model. The alternating direction multiplier method is then used for each center frequency. Recursive correction is performed. Specifically, the sum of the energy bandwidths of all sub-signals is monitored in real time. When this sum decreases and meets a preset condition... The iteration stops when the time is reached. This yields the updated center frequency vector. ,in, A positive integer, representing the total number of center frequencies. They represent the 1st, 2nd, and 3rd respectively. The and the first The updated center frequencies are obtained by applying multiple rounds of variational iteration constraints to the initial center frequency of the time-frequency signal using the alternating direction multiplier method, until the sum of the energy bandwidths continuously decreases and meets a preset condition. When the convergence criterion is met, the final steady-state frequency value output is used to extract the time-frequency signal sequence using the updated parameters. Multiple optimized sub-components were extracted from the sample.

[0066] Step 305: Utilize optimized sub-components to extract components from the time-frequency signal, obtaining multiple target sub-signals, and use the target sub-signals as intrinsic mode components.

[0067] In this step, the target sub-signal refers to a narrowband oscillation component that has independent physical meaning and minimal interference.

[0068] In this embodiment, a Wiener filter is first constructed based on the frequency characteristics of optimized sub-components. This is applied to the time-frequency signal sequence. Perform component extraction. Decouple the original signal into multiple target sub-signals reflecting different physical meanings. For example, for tank No. 1 at gas station A, extract the component representing the trend of tank cross-sectional area change and the component representing random noise fluctuations. Define these target sub-signals as intrinsic mode component vectors.

[0069] Step 104: Extract the low-frequency intrinsic mode components from the intrinsic mode components through the energy centroid and superimpose and reconstruct them to obtain the target intrinsic sequence representing the relationship between the height and volume of the earth tank.

[0070] In this step, the energy centroid refers to the signal frequency center point determined by the sum of the products of the frequency value and its corresponding amplitude weight. Low-frequency intrinsic mode components refer to signal components whose energy centroid frequency is below a preset threshold. The target intrinsic sequence refers to the sequence of height and volume correspondences that has eliminated random noise interference and can accurately represent the geometry of the underground tank.

[0071] Step 401: Based on the amplitude information in each intrinsic mode component, the frequency values ​​in each intrinsic mode component are weighted and summed to obtain the energy centroid frequency of each intrinsic mode component.

[0072] In this step, the energy centroid frequency refers to the frequency value that represents the center position of the component, obtained by weighted averaging of the spectral energy distribution of each intrinsic mode component.

[0073] In this embodiment, the intrinsic mode component vectors obtained through variational mode decomposition are first extracted. ,in, These represent the 1st, 2nd, and 3rd results obtained from the decomposition, respectively. The and the first Each intrinsic mode component A positive integer, representing the total number of intrinsic mode components, followed by each component... Obtain its magnitude vector and the corresponding frequency vector ,in, They represent the first The 1st, 2nd, and 3rd eigenmode components The and the first The amplitude of each frequency node; They represent the first The 1st, 2nd, and 3rd eigenmode components The and the first The frequency values ​​corresponding to each frequency node are processed using weighted averages and techniques. Specifically, as shown in formula (1): ; Calculate the first The energy center frequency of each component A positive integer, representing the first... The total number of discrete frequency nodes contained in each intrinsic mode component; A positive integer representing the index number of the frequency node. , This indicates the first element in the component. The amplitude corresponding to each frequency node; Indicates the first The frequency values ​​corresponding to each frequency node. For example, for intrinsic mode components. The energy centroid frequency is determined by the ratio of the sum of the products of its amplitude distribution and frequency distribution to the total sum of amplitudes. .

[0074] Step 402: Determine the intrinsic mode components corresponding to the energy centroid frequencies that are less than the preset frequency threshold as low-frequency intrinsic mode components.

[0075] In this step, the preset frequency threshold refers to the frequency limit set according to the evolution law of the earthwork's geometry with height, used to distinguish characteristic signals from random interference. Low-frequency intrinsic mode components refer to signal components whose energy centroid frequency is in a lower frequency band and can represent the trend of changes in the earthwork's macroscopic cross-sectional area.

[0076] In this embodiment, a preset frequency threshold is first obtained. Next, the energy-centroid frequency of each eigenmode component is... Compare each value with the threshold. If the logical condition is met... Less than If so, it is determined to be a low-frequency intrinsic mode component.

[0077] For example, when processing the data of tank No. 1 at gas station A, if the calculated energy centroid frequency sequence is... ,in, These represent the energy centroid frequencies of the first, second, and third eigenmode components obtained from the decomposition, respectively, and comparisons reveal that only... Less than the preset frequency threshold Then the components Low-frequency intrinsic mode components are identified and retained, while high-frequency components are... and As noise removal.

[0078] Step 403: Using the height nodes on the liquid level coordinate axis as the horizontal coordinate index, extract the volume values ​​corresponding to the low-frequency intrinsic mode components at each height node and accumulate them to obtain the composite volume at each height node. Then, bind the composite volume to the corresponding height node to obtain the target intrinsic sequence.

[0079] In this step, the synthetic volume refers to the numerical value reconstructed by superimposing the volumetric characteristic values ​​corresponding to all selected low-frequency intrinsic mode components at the same height node location. The target intrinsic sequence refers to the data sequence that reflects the essential correspondence between the height and volume of the underground tank, generated by binding the synthetic volume sequence to the height node sequence.

[0080] In this embodiment of the application, the liquid level coordinate axis is first used as an example. The sequence of height nodes on This is the x-axis index. These represent the 1st, 2nd, and 3rd height nodes in the sequence, respectively. The and the first Each height node. Then at each height node Extract the volume values ​​carried by all determined low-frequency intrinsic mode components. The process is handled using overlay and reconstruction techniques. This is achieved through formulas. The composite volume at each height node is calculated.

[0081] For example, during the calibration process of tank No. 1, if the selected component is... Then extract its height The volume component at each height node. The composite volume at all height nodes. With the corresponding height node Binding is performed. The final result is a target intrinsic sequence composed of height and volume data pairs. ,in, They represent the 1st, 2nd, and 3rd elements in the target intrinsic sequence, respectively. The and the first A data pair consisting of a height node and a composite volume. Represents the target intrinsic sequence; is a positive integer, representing the total number of height nodes included in the target intrinsic sequence.

[0082] Step 105: Determine the theoretical volume sequence based on the pre-built earth tank model, and construct the residual function based on the difference between the target intrinsic sequence and the theoretical volume sequence.

[0083] In this step, the pre-constructed tank model refers to the height-volume functional relationship established based on an ideal geometry. The theoretical volume sequence refers to the set of height-corresponding volumes obtained through mathematical calculations using the tank model under specific geometric parameters. The residual function is the objective function used to evaluate the goodness of fit of the model, with geometric parameters as independent variables.

[0084] Step 501: Use the ground tank model to perform volume inversion calculations at each height node to obtain the theoretical volume sequence.

[0085] A ground tank model refers to a functional relationship between height and volume established based on an ideal geometric shape. Specifically, the ground tank model adopts a horizontal cylindrical geometric structure, and its volume-liquid level relationship can be expressed as: ,in, This represents the total theoretical volume of the tank model at a specific liquid level. This represents the volume of the cylindrical section when tilted. This represents the volume of the end caps. The calculation considers the equivalent change in liquid level height under the tilt angle. The theoretical volume sequence refers to the set of volumes corresponding to heights obtained through mathematical calculations using a ground tank model under specific geometric parameters.

[0086] In this embodiment of the application, the initial geometric parameter vector is extracted. Initial geometric parameters refer to the set of geometric dimensions pre-set according to the design specifications or historical records of the underground tank before the start of the tank alignment operation. These parameters may include the nominal diameter and nominal length of the tank. Indicates the nominal diameter of the earthen tank; This represents the nominal length of the tank. The mathematical formulas for the pre-built tank model are used at each height node. Volume inversion calculations are performed at the location. The theoretical volume values ​​are obtained by substituting the height nodes as independent variables into the geometric equations.

[0087] For example, in the scenario of tank No. 1 at gas station A, the parameter vector Input model, for the sequence of height nodes By solving point by point, the results are finally integrated to generate a theoretical volume sequence composed of the volumes of each node. ,in, These represent the 1st, 2nd, and 3rd volumes in the theoretical volume sequence, respectively. The and the first The theoretical volume value of each height node.

[0088] Step 502: Calculate the numerical distance between each theoretical volume in the theoretical volume sequence and the corresponding synthetic volume in the target intrinsic sequence to obtain the residual set.

[0089] In this embodiment of the application, the theoretical volume sequence is first... With the target intrinsic sequence Perform height-dimensional alignment. Extract the theoretical volume corresponding to each height node. and synthesis volume Next, the absolute value of the difference between the two is calculated, representing the deviation at that point. Specifically, in the calibration process of tank No. 1, this is done by comparing the theoretical volume sequence. and target intrinsic sequence Using the formula The numerical spacing at each sampling node is obtained, and finally, the spacing values ​​of all nodes are combined to generate a residual set vector composed of error components. ,in, They represent the 1st, 2nd, and 3rd elements in the residual set vector, respectively. The and the first Error components at each sampling node.

[0090] Step 503: Calculate the sum of squares of all numerical intervals in the residual set to obtain the sum. Based on the mapping relationship between the sum and the initial geometric parameters, construct the residual function.

[0091] In this step, the summation refers to the scalar value reflecting the overall fitting error, obtained by accumulating the sums of squares. The mapping relationship refers to the corresponding logic of the summation as the geometric parameters change.

[0092] In this embodiment of the application, the residual set vector is first... Each numerical spacing Perform the squaring operation. Use the summation technique to calculate the sum of all squared values. As shown in formula (2): ; Get the sum .in, Indicates the parameters of the earthen tank model and height Theoretical output value below; This represents the composite volume value corresponding to the height node in the target intrinsic sequence. Based on cumulative summation. With initial geometric parameters The logical dependencies between them are used to construct residual functions.

[0093] For example, in the calculation process of gas station A, parameters are established. This is a mathematical expression with independent variables and a cumulative sum as the dependent variable. This function describes the impact of the evolution of model parameters on the overall error.

[0094] Step 106: Iteratively optimize the initial geometric parameters of the earth tank model using the fractional gradient optimization differential operator to determine the target geometric parameters that minimize the residual function, and map the target geometric parameters to the earth tank model. Generate a correction volume table through piecewise numerical integration across the entire range.

[0095] In this step, fractional gradient optimization refers to an optimization strategy that uses non-integer derivative information to search for the extremum of the objective function. This application preferably uses the Caputo fractional differential operator, which possesses global memory and nonlocality characteristics when dealing with nonlinear geometric deviation optimization. A differential operator is a mathematical tool used to calculate the fractional derivative of a function. Initial geometric parameters refer to the original parameter estimates when the model begins iteration. Initial geometric parameters are the geometric dimensions pre-set according to the tank's design specifications before iterative optimization, and may include the tank's diameter, cylinder length, installation tilt angle, and end cap shape coefficient. Target geometric parameters refer to the optimal geometric parameter solution that makes the residual function reach its global minimum. Full-range piecewise numerical integration refers to a numerical calculation method that subdivides the tank model into multiple small units according to height and accumulates the volume piecewise. The corrected volume table refers to a numerical mapping table generated after accurate calculation using the optimized geometric parameters, reflecting the correspondence between the actual liquid level height and the oil volume in the tank.

[0096] like Figure 3 As shown, Figure 3 This is a schematic flowchart illustrating the method for generating a correction volume table provided in an embodiment of this application.

[0097] Step 601: Calculate the fractional derivatives of the initial geometric parameters using differential operators to determine the direction and step size of parameter adjustment.

[0098] In this step, the parameter adjustment direction refers to the geometric evolution dimension in the parameter space that indicates the fastest decrease in the residual function value. The parameter adjustment step size refers to the magnitude of the change in geometric parameters during each round of optimization calculation.

[0099] In this embodiment, initial geometric parameters are first extracted. Fractional derivative operations are performed on the residual function using a differential operator. The direction of parameter adjustment is determined by analyzing the obtained gradient vector. The parameter adjustment step size is determined by combining a preset learning rate weight. Specifically, in the parameter optimization process of tank No. 1, the current initial geometric parameter vector is... and preset score levels Fractional order The range of values ​​is set to Preferably 0.8 or 1.2, adjusted by... It can effectively control the damping characteristics of gradient descent, avoiding the problem of traditional first-order gradient algorithms easily getting trapped in local optima. The adjustment direction vector for parameter evolution is obtained through differential operations. and adjusting step size ,in These represent the nominal diameters of the earthen tanks. and nominal length The adjustment direction components in the parameter space are specifically valued by the gradients generated by solving the partial derivatives of the aforementioned residual function with respect to the nominal diameter D and nominal length L using fractional differential operators; the adjustment step size... The preset learning rate weights.

[0100] Step 602: Based on the parameter adjustment direction and parameter adjustment step size, update the initial geometric parameters to obtain the updated initial geometric parameters. Substitute the updated initial geometric parameters into the residual function to obtain the residual value. If the residual value is greater than the preset convergence threshold, return to use the differential operator to calculate the fractional derivative of the initial geometric parameters until the residual value is less than or equal to the preset convergence threshold. Then, use the updated initial geometric parameters as the target geometric parameters.

[0101] In this step, the preset convergence threshold is a small positive scalar value used to determine whether the iterative optimization process has reached the predetermined accuracy requirement and allows the calculation to stop. The target geometric parameters refer to the final determined actual structural dimensions of the tank when the convergence conditions are met.

[0102] In this embodiment, the initial geometric parameters are first updated based on the parameter adjustment direction and parameter adjustment step size to obtain the updated initial geometric parameters. The update logic can be shown in formula (3): ; in, Indicates the first The geometric parameter vector of the earth tank at the next iteration; Indicates the first The updated geodetic parameter vector of the earthen tank in the next iteration; A positive integer representing the number of iterations; Indicates the order of a fractional differential operator; The step size for parameter adjustment during the iteration process is the learning rate. This indicates that when the parameter is The residual function at that time.

[0103] The residual value is then substituted into the residual function to obtain the current residual value. This value is then compared to a preset convergence threshold. If the residual value is greater than the preset convergence threshold, the process returns to using a differential operator to calculate the fractional derivative of the initial geometric parameters. This continues until the residual value is less than or equal to the preset convergence threshold. Finally, the updated parameters are used as the target geometric parameters.

[0104] For example, in the scenario of gas station A, the residual value is converged to a preset convergence threshold through iterative iteration. Within the range. The result extracted at this point is the target geometric parameter vector. .in, This represents the target geometric parameter vector that makes the residual function converge; This represents the target nominal diameter of the tank, determined ultimately through iterative optimization. This represents the target nominal length of the ground tank, which is finally determined through iterative optimization.

[0105] Step 603: Using the total range of the liquid level coordinate axis as a reference, the tank model is cut into micro-elements in the height dimension using the target geometric parameters to obtain multiple volume micro-element values.

[0106] In this step, the total range of the liquid level coordinate axis refers to the complete vertical height range covered by the tank, from the lowest liquid level sampling point at the bottom to the highest safe liquid level sampling point at the top. The volume element value refers to the internal space volume value of the model within each small height range.

[0107] In this embodiment, the total range of the liquid level coordinate axis is first obtained. An accurate tank model is then constructed using the determined target geometric parameters. Next, the model is divided into infinitesimal elements along its height dimension. Specifically, tank No. 1 is divided along the height direction at fixed steps. It is divided into multiple horizontal thin layers. The integration step size is as follows: The selection of parameters balances computational efficiency and full-range accuracy, ensuring that the volumetric meter achieves millimeter-level precision. Microscopic geometric parameters are obtained through piecewise numerical integration. The optimization results are transformed into a macroscopic volume mapping, which, compared to traditional polynomial fitting, can capture subtle local deformations of the earthwork tank. The volume corresponding to each layer is calculated using geometric formulas. Finally, a sequence of volume element values ​​consisting of all components across the entire height range is generated. ,in, These represent the first, second, and third elements after the high-dimensional infinitesimal element is cut. Each of the volume element values ​​is the target nominal diameter obtained through optimization. and target nominal length After substituting the geometric parameters into the earthen tank model, the theoretical volume of a single layer is accurately calculated within the corresponding horizontal thin-layer integration step, combined with the derivation formula of the geometric boundary of the horizontal cylinder. is a positive integer, representing the total number of volumetric elements after the height-dimensional element is cut.

[0108] Step 604: Accumulate and sum the volume element values ​​segment by segment according to the ascending height order to determine the cumulative volume value corresponding to each height node. Map the cumulative volume value to the corresponding height node to obtain the corrected volume table.

[0109] In this step, the cumulative volume value refers to the total volume inside the tank, stacked from the bottom up to the current height node.

[0110] In this embodiment, the initial volume is first set to zero. The volume element values ​​are then sorted in ascending order of height. Perform segment-by-segment cumulative summation. Use addition to determine the cumulative volume value corresponding to each height node. Finally, map each cumulative volume value to its corresponding height node. For example, by using the height node sequence... By binding the cumulative volume numerical sequence, the corrected volume table of tank No. 1 at gas station A is obtained. This table can be represented by a mapping matrix: ;

[0111] This application embodiment obtains refueling details, real-time liquid level data, and unloading data of gas station tanks within a preset time period, ensuring the comprehensiveness of the analysis sample. It achieves a high degree of alignment between sales data and liquid level changes, eliminating interference fluctuations during non-trading periods, and improving the consistency and reliability of the original data. It overcomes the limitation of traditional optimization algorithms that easily get trapped in local optima, achieving global and accurate capture of complex nonlinear deviations in the tanks, and ensures that the final generated volume table has extremely high correction accuracy across the entire range through piecewise numerical integration.

[0112] Figure 4 This is a schematic diagram of a specific implementation of the calibration system for the gas station tank volume table provided in this application embodiment, with reference to... Figure 4 The system may include: The acquisition module 21 is used to acquire the refueling details, real-time liquid level data and unloading data of the gas station's underground tank within a preset time period. The unloading data includes the unloading volume and unloading height. The determination module 22 is used to extract multiple sample points, including transaction volume and actual liquid level change height, from refueling details and real-time liquid level data through self-matching, and to determine the volume compensation amount based on unloading data, refueling details during unloading, and volume table. Module 23 is used to construct time-frequency signals based on sample points and volume compensation amount, using the mapping operator in the multidimensional signal fusion analysis algorithm, and to perform non-recursive decomposition on the time-frequency signals using variational mode decomposition to obtain multiple intrinsic mode components. Module 23 is also used to extract low-frequency intrinsic mode components from intrinsic mode components through energy centroid and superimpose and reconstruct them to obtain the target intrinsic sequence representing the relationship between the height and volume of the earth tank; Module 23 is also used to determine the theoretical volume sequence based on the pre-built earth tank model and to construct the residual function based on the difference between the target intrinsic sequence and the theoretical volume sequence. The determination module 22 is also used to iteratively optimize the initial geometric parameters of the earth tank model using a fractional gradient optimization differential operator to determine the target geometric parameters that minimize the residual function, and to map the target geometric parameters to the earth tank model, and generate a correction volume table through full-range piecewise numerical integration.

[0113] The calibration system for the gas station tank volume meter of this application embodiment is used to implement the aforementioned calibration method for the gas station tank volume meter. Therefore, the specific implementation of the calibration system for the gas station tank volume meter can be found in the embodiment section of the calibration method for the gas station tank volume meter above. The specific implementation can be referred to the description of the corresponding embodiments, which will not be repeated here.

[0114] Figure 5 A schematic diagram of the hardware structure of the electronic device provided in an embodiment of this application is shown.

[0115] This application also provides an electronic device, comprising: a memory for storing a computer program; and a processor for executing the computer program to implement the steps of the above-described method for correcting the volume table of a gas station tank.

[0116] The electronic device may include a processor 510 and a memory 520 storing computer program instructions.

[0117] Specifically, the processor 510 may include a central processing unit (CPU), an application-specific integrated circuit (ASIC), or one or more integrated circuits that can be configured to implement the embodiments of this application.

[0118] Memory 520 may include mass storage for data or instructions. For example, and not limitingly, memory 520 may include a hard disk drive (HDD), floppy disk drive, flash memory, optical disk, magneto-optical disk, magnetic tape, or Universal Serial Bus (USB) drive, or a combination of two or more of these. Where appropriate, memory 520 may include removable or non-removable (or fixed) media. Where appropriate, memory 520 may be internal or external to the integrated gateway disaster recovery device. In a particular embodiment, memory 520 is non-volatile solid-state memory.

[0119] Memory may include read-only memory (ROM), random access memory (RAM), disk storage media devices, optical storage media devices, flash memory devices, and electrical, optical, or other physical / tangible memory storage devices. Therefore, typically, memory includes one or more tangible (non-transitory) computer-readable storage media (e.g., memory devices) encoded with software including computer-executable instructions, and when the software is executed (e.g., by one or more processors), it is operable to perform the operations described with reference to the method according to the first aspect of this disclosure.

[0120] The processor 510 reads and executes computer program instructions stored in the memory 520 to implement the calibration method for the tank volume table of any gas station in the above embodiments.

[0121] In one example, the electronic device may also include a communication interface 530 and a bus 540. Wherein, such as Figure 5 As shown, the processor 510, memory 520, and communication interface 530 are connected through bus 540 and complete communication with each other.

[0122] The communication interface 530 is mainly used to realize communication between various modules, devices, units and / or equipment in the embodiments of this application.

[0123] Bus 540 includes hardware, software, or both, that couples components of an online data traffic metering device together. For example, and not limitingly, the bus may include an Accelerated Graphics Port (AGP) or other graphics bus, an Enhanced Industry Standard Architecture (EISA) bus, a Front Side Bus (FSB), HyperTransport (HT) interconnect, an Industry Standard Architecture (ISA) bus, an Infinite Bandwidth Interconnect, a Low Pin Count (LPC) bus, a memory bus, a Microchannel Architecture (MCA) bus, a Peripheral Component Interconnect (PCI) bus, a PCI-Express (PCI-X) bus, a Serial Advanced Technology Attachment (SATA) bus, a Video Electronics Standards Association Local (VLB) bus, or other suitable buses, or combinations of two or more of these. Where appropriate, bus 540 may include one or more buses. Although specific buses are described and illustrated in embodiments of this application, any suitable bus or interconnect is contemplated herein.

[0124] This application also provides a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the steps of the above-described method for correcting the volume table of a gas station tank.

[0125] In one exemplary embodiment, the aforementioned computer-readable storage medium may include, but is not limited to, various media capable of storing computer programs, such as USB flash drives, read-only memory, random access memory, portable hard drives, magnetic disks, or optical disks.

[0126] Embodiments of the present invention also provide a computer program product, which includes a computer program that, when executed by a processor, implements the steps in the above-described method for correcting the volume table of a gas station tank.

[0127] Those skilled in the art will further recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of both. To clearly illustrate the interchangeability of hardware and software, the components and steps of the various examples have been generally described in terms of functionality in the foregoing description. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementations should not be considered beyond the scope of this invention.

[0128] The above provides a detailed description of the calibration method and system for the gas station tank volume table provided in this application. Specific examples have been used to illustrate the principles and implementation methods of this application. The descriptions of the embodiments above are merely for the purpose of helping to understand the method and its core ideas. It should be noted that those skilled in the art can make various improvements and modifications to this application without departing from its principles, and these improvements and modifications also fall within the protection scope of this application.

Claims

1. A method for calibrating the volume gauge of a gas station's underground tank, characterized in that, include: The system acquires refueling details, real-time liquid level data, and unloading data of the gas station's underground tank within a preset time period. The unloading data includes the unloading volume and unloading height. Multiple sample points, including transaction volume and actual liquid level change height, are extracted from the refueling details and the real-time liquid level data through self-matching, and the volume compensation amount is determined based on the unloading data, the refueling details and the volume table during the unloading period. Based on the sample points and the volume compensation amount, a time-frequency signal is constructed using the mapping operator in the multidimensional signal fusion analysis algorithm, and the time-frequency signal is non-recursively decomposed using variational mode decomposition to obtain multiple intrinsic mode components. The low-frequency intrinsic mode components are extracted from the intrinsic mode components by the energy centroid and then superimposed and reconstructed to obtain the target intrinsic sequence representing the relationship between the height and volume of the earth tank. The theoretical volume sequence is determined based on the pre-constructed earth tank model, and a residual function is constructed based on the difference between the target intrinsic sequence and the theoretical volume sequence. The initial geometric parameters of the earth tank model are iteratively optimized using a fractional gradient optimization differential operator to determine the target geometric parameters that minimize the residual function. The target geometric parameters are then mapped to the earth tank model, and a correction volume table is generated through piecewise numerical integration across the entire range.

2. The method according to claim 1, characterized in that, The refueling details include the transaction time and transaction volume, and the real-time liquid level data includes the real-time sampling height and timestamp. Multiple sample points, including the transaction volume and the actual liquid level change height, are extracted from the refueling details and the real-time liquid level data through self-matching, including: Based on the transaction time, multiple transaction time periods corresponding to each transaction volume are determined; Based on the transaction time period, retrieve the corresponding timestamp, and extract the start sampling height and end sampling height corresponding to each transaction time period from the real-time sampling height; Subtract the corresponding starting sampling height from each termination sampling height to obtain the actual liquid level change height for each trading period. Using the transaction time period as an index, each actual liquid level change height is associated with the corresponding transaction volume to obtain multiple sample points.

3. The method according to claim 1, characterized in that, The volume compensation amount is determined based on the unloading data, the refueling details during the unloading period, and the volume table, including: The total sales volume is obtained by summing the volumes of all transactions within the unloading period in the refueling details. Subtract the total sales volume from the oil unloading volume on the voucher to obtain the actual oil volume. Using the volume table, look up the first volume corresponding to the oil unloading start height and the second volume corresponding to the oil unloading end height, and subtract the first volume from the second volume to obtain the theoretical oil inlet volume; The volume compensation amount is obtained by subtracting the theoretical oil intake volume from the actual oil volume.

4. The method according to claim 2, characterized in that, Based on the sample points and the volume compensation amount, a time-frequency signal is constructed using the mapping operator in the multidimensional signal fusion analysis algorithm. The time-frequency signal is then non-recursively decomposed using variational mode decomposition to obtain multiple intrinsic mode components, including: Based on the real-time sampling height and the unloading height, a liquid level coordinate axis is constructed. The mapping operator is used to map the transaction volume and the volume compensation amount corresponding to each sample point to the liquid level coordinate axis, respectively, to obtain multiple fusion points. Based on the height value of each fusion point, all fusion points are sorted in ascending order to obtain an ordered point set. The volume values ​​in the ordered point set are then interpolated with equal step size to obtain the time-frequency signal. The time-frequency signal is subjected to spectral analysis to obtain a spectral distribution, which includes multiple sub-signals and a center frequency corresponding to each sub-signal. Each sub-signal includes amplitude information and frequency value. Based on the diffusion degree of each sub-signal relative to the corresponding center frequency, the energy bandwidth of each sub-signal is calculated, and each center frequency is iteratively adjusted until the sum of the energy bandwidths of all sub-signals meets the preset condition, thereby obtaining the updated center frequency. Based on the updated center frequency, multiple optimized sub-components are extracted from the time-frequency signal. The optimized sub-components are used to extract components from the time-frequency signal to obtain multiple target sub-signals, which are then used as intrinsic mode components.

5. The method according to claim 4, characterized in that, Low-frequency intrinsic mode components are extracted from the intrinsic mode components using the energy centroid and then superimposed and reconstructed to obtain the target intrinsic sequence representing the relationship between the height and volume of the earth tank, including: Based on the amplitude information in each intrinsic mode component, the frequency values ​​in each intrinsic mode component are weighted and summed to obtain the energy centroid frequency of each intrinsic mode component; The intrinsic mode components corresponding to the energy centroid frequency that is less than a preset frequency threshold are determined as low-frequency intrinsic mode components. Using the height nodes on the liquid level coordinate axis as the horizontal coordinate index, the volume values ​​corresponding to the low-frequency intrinsic mode components are extracted at each height node and accumulated to obtain the composite volume at each height node. The composite volume is then bound to the corresponding height node to obtain the target intrinsic sequence.

6. The method according to claim 5, characterized in that, The theoretical volume sequence is determined based on a pre-constructed earthen tank model. A residual function is constructed based on the difference between the target intrinsic sequence and the theoretical volume sequence, including: The theoretical volume sequence is obtained by performing volume inversion calculations at each height node using the aforementioned tank model. Calculate the numerical distance between each theoretical volume in the theoretical volume sequence and the corresponding synthetic volume in the target intrinsic sequence to obtain the residual set; The sum of squares of all numerical intervals in the residual set is calculated to obtain the cumulative sum. Based on the mapping relationship between the cumulative sum and the initial geometric parameters, the residual function is constructed.

7. The method according to claim 6, characterized in that, The initial geometric parameters of the earthen tank model are iteratively optimized using a fractional gradient optimization differential operator to determine the target geometric parameters that minimize the residual function. These target geometric parameters are then mapped to the earthen tank model. A corrected volume table is generated through piecewise numerical integration across the entire range, including: The fractional derivative of the initial geometric parameters is calculated using the differential operator to determine the parameter adjustment direction and parameter adjustment step size. Based on the parameter adjustment direction and the parameter adjustment step size, the initial geometric parameters are updated to obtain the updated initial geometric parameters. The updated initial geometric parameters are then substituted into the residual function to obtain the residual value. If the residual value is greater than a preset convergence threshold, the fractional derivative of the initial geometric parameters is calculated using the differential operator until the residual value is less than or equal to the preset convergence threshold. The updated initial geometric parameters are then used as the target geometric parameters. Using the total range of the liquid level coordinate axis as a reference, the ground tank model is divided into height-dimensional micro-elements using the target geometric parameters to obtain multiple volume micro-element values; The volume element values ​​are accumulated and summed segment by segment according to the ascending height order to determine the cumulative volume value corresponding to each height node. The cumulative volume value is then mapped to the corresponding height node to obtain the corrected volume table.

8. A calibration system for the volume gauge of a gas station's underground tank, characterized in that, include: The acquisition module is used to acquire refueling details, real-time liquid level data and unloading data of the gas station's underground tank within a preset time period. The unloading data includes the unloading volume and unloading height. The determination module is used to extract multiple sample points, including transaction volume and actual liquid level change height, from the refueling details and the real-time liquid level data through self-matching, and to determine the volume compensation amount based on the unloading data, the refueling details and the volume table during the unloading period. The construction module is used to construct a time-frequency signal based on the sample points and the volume compensation amount using the mapping operator in the multidimensional signal fusion analysis algorithm, and to perform non-recursive decomposition on the time-frequency signal using variational mode decomposition to obtain multiple intrinsic mode components. The construction module is also used to extract low-frequency intrinsic mode components from the intrinsic mode components through the energy centroid and superimpose and reconstruct them to obtain a target intrinsic sequence representing the relationship between the height and volume of the earth tank; The construction module is also used to determine the theoretical volume sequence based on the pre-built earth tank model, and to construct a residual function based on the difference between the target intrinsic sequence and the theoretical volume sequence; The determination module is also used to iteratively optimize the initial geometric parameters of the earth tank model using a fractional gradient optimization differential operator to determine the target geometric parameters that minimize the residual function, and to map the target geometric parameters to the earth tank model, and generate a correction volume table through full-range piecewise numerical integration.

9. An electronic device, characterized in that, include: Memory, used to store computer programs; A processor, configured to execute the computer program to implement the steps of the method for correcting the volume table of the gas station tank as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that, when executed by a processor, enables the correction method for the gas station tank volume table as described in any one of claims 1 to 7.

Images