An underground pipeline data desensitization method and device, electronic equipment and storage medium

Abstract

The application provides an underground pipeline data desensitization method and device, electronic equipment and a storage medium. The underground pipeline control point is acquired, the coordinates of the underground pipeline control point are taken as the training behavior of input data, and a neural network transformation model is established. Then, the desensitization data is acquired, the desensitization data is desensitized by the neural network transformation model, intermediate data is acquired, a disturbance amount is applied to the intermediate data, the model coefficients of the underground pipeline data desensitization model are acquired by calculating the intermediate data and the intermediate data disturbance amount through Chebyshev polynomials, the intermediate data is input into the underground pipeline data desensitization model, and final desensitization data is acquired. The original underground pipeline data is subjected to multi-stage transformation and multi-stage desensitization by using two different desensitization models, so that the anti-attack ability is stronger and the complexity is higher. Meanwhile, the desensitized result is controlled by using a desensitization index, so that the data security is improved on the basis of ensuring the data availability.

Description

Technical Field

[0001] This application relates to the field of data processing, and in particular to a method, apparatus, electronic device, and storage medium for desensitizing underground pipeline data. Background Technology

[0002] With the acceleration of urbanization in my country, urban construction projects require frequent road excavation. However, without sufficient information on the current status of underground pipelines, construction companies are prone to pipeline rupture accidents. Achieving information sharing and providing data publishing services with planning, construction, and operation departments or units while ensuring information security and confidentiality remains a significant challenge.

[0003] Current technologies for data anonymization typically employ methods such as similarity transformation, affine transformation, projective transformation, and radial basis function models. However, these traditional methods leave the anonymized data vulnerable to cracking and theft, resulting in low usability and failing to simultaneously meet the practical requirements of confidentiality and usability. Currently, no data anonymization method has been developed specifically for underground pipelines, where high confidentiality is required while maintaining data accuracy. Summary of the Invention

[0004] This invention provides a method, apparatus, electronic device, and storage medium for desensitizing underground pipeline data. It performs spatial desensitization on underground pipeline data through a multi-level desensitization transformation method using neural network transformation and Chebyshev model, which ensures both strong confidentiality and data accuracy.

[0005] In a first aspect, the present invention provides a method for desensitizing underground pipeline data, comprising the following steps:

[0006] Obtain underground pipeline control points, and establish a neural network transformation model based on the training behavior of the coordinates of the underground pipeline control points as input data;

[0007] Obtain the data to be de-identified, and perform de-identification processing on the data to be de-identified through the neural network transformation model to obtain intermediate data;

[0008] A perturbation amount is applied to the intermediate data, and the model coefficients of the underground pipeline data desensitization model are obtained by Chebyshev polynomial calculation based on the intermediate data and the perturbation amount.

[0009] The intermediate data is input into the underground pipeline data desensitization model to obtain the final desensitized data.

[0010] Furthermore, after obtaining the intermediate data, the process also includes:

[0011] Control point Cpoint(sX) is obtained based on the selection of intermediate data.i ,sY i The control points are then subjected to coordinate normalization.

[0012] Apply a disturbance (ΔX) to the control point i ,ΔY i ), obtain the desensitized control point (tX) i ,tY i );

[0013] Based on the coordinates of the control points and the coordinates of the desensitized control points, the coefficients of the Chebyshev polynomial matrix are calculated using the least squares method.

[0014] Based on the Chebyshev polynomial matrix coefficients and the Chebyshev expression for the desensitized control point coordinates, a data desensitization model for underground pipelines is established. The Chebyshev expression for the desensitized control point coordinates is:

[0015]

[0016] Among them, CX ij and CY ij ΔX i with ΔY i Chebyshev polynomial coefficients, T i (sX ig ) and T j (sY ig ) are the i-th and j-th Chebyshev polynomials with the x and y coordinates as independent variables after normalization, respectively, and m and n are the cutoff orders of the Chebyshev polynomials.

[0017] Furthermore, obtaining the model coefficients of the anonymized underground pipeline data model further includes:

[0018] The coordinates of the control points are used as the input data of the underground pipeline data desensitization model, and the disturbance of the control points is used as the output data of the underground pipeline data desensitization model to obtain the model coefficients to be adjusted.

[0019] Update the underground pipeline data desensitization model based on the model coefficients to be adjusted;

[0020] Sample points are obtained based on the intermediate data, and desensitized sample data is obtained through the updated underground pipeline data desensitization model.

[0021] Based on the desensitized sample data, the desensitization error is calculated. If the difference between the desensitization error and the preset desensitization index exceeds the preset first threshold, the disturbance amount is iteratively processed.

[0022] When the difference between the desensitization error and the preset desensitization index is less than the preset first threshold, the iteration of the disturbance amount is stopped, and the coefficients of the current underground pipeline data desensitization model are determined as the coefficients of the underground pipeline data desensitization model used to calculate the final desensitized data.

[0023] Furthermore, the establishment of the neural network transformation model also includes:

[0024] Based on the coordinates of the underground pipeline control points, obtain neural network sample points (sX). i ,sY i );

[0025] Based on the coordinates of the sample points in the aforementioned neural network, the transformed neural network sample points (stX) are obtained through the aforementioned neural network transformation model. i ,stY i );

[0026] Based on the transformed neural network sample points (stX) i ,stY i The transformation error is calculated. If the difference between the transformation error and a preset transformation index exceeds a preset second threshold, the expansion constant in the neural network transformation model is iteratively processed. The formula for calculating the transformation error is:

[0027]

[0028] The iterative formula for the extended constant is:

[0029]

[0030] Where, d sample The transformation error is denoted as θ, the expansion constant is σ1, and the preset transformation index is σ1.

[0031] When the transformation error d sample When the difference between the value of the expansion constant θ and the preset transformation index σ1 is less than the preset second threshold, the iteration of the expansion constant θ is stopped, and the neural network transformation model is updated according to the weights of the current neural network transformation model.

[0032] Furthermore, before obtaining the intermediate data, the process also includes:

[0033] Based on the adjacent underground pipeline control points, obtain the coordinates of the interpolation point;

[0034] The neural network transformation model is used to obtain the transformed coordinates of the adjacent underground pipeline control points and the interpolation points.

[0035] A preset point insertion threshold is set. If the distance between the transformed interpolated point coordinates and the straight line where the adjacent underground pipeline control point is located exceeds the point insertion threshold, then the interpolated point is added to the underground pipeline control point.

[0036] Furthermore, after obtaining the final de-identified data, the process also includes:

[0037] A key file is stored, which includes the control point, the disturbance amount, and the desensitized sample data.

[0038] Furthermore, the iterative processing of the disturbance quantity specifically includes:

[0039] The formula for calculating the desensitization error is as follows:

[0040]

[0041] Wherein, RMSE is the desensitization error, and sX is the desensitization error. i ,sY i Let tX be the x and y coordinates of the sample point. i ,tY i These are the x and y coordinates of the desensitized sample points;

[0042] The disturbance amount (ΔX) i ,ΔY i The iterative formula for ) is:

[0043]

[0044] Wherein, σ2 is the preset desensitization index;

[0045] When RMSE < σ², increase the disturbance (ΔX) i ,ΔY i If RMSE > σ², reduce the disturbance (ΔX). i ,ΔY i ).

[0046] Secondly, the present invention also provides an underground pipeline data desensitization device, comprising:

[0047] The neural network transformation model establishment module is used to obtain underground pipeline control points and establish a neural network transformation model based on the coordinates of the underground pipeline control points as input data.

[0048] The intermediate data acquisition module is used to acquire the data to be desensitized, and to desensitize the data to be desensitized through the neural network transformation model to acquire intermediate data;

[0049] The model coefficient acquisition module is used to apply a perturbation amount to the intermediate data, and calculate the model coefficients of the underground pipeline data desensitization model based on the intermediate data and the perturbation amount using Chebyshev polynomials.

[0050] The desensitized data acquisition module is used to input the intermediate data into the underground pipeline data desensitization model to obtain the final desensitized data.

[0051] Thirdly, embodiments of this application provide an electronic device, including:

[0052] At least one memory and at least one processor;

[0053] The memory is used to store one or more programs;

[0054] When the one or more programs are executed by the at least one processor, the at least one processor performs the steps of the underground pipeline data desensitization method as described in the first aspect.

[0055] Fourthly, embodiments of this application provide a computer-readable storage medium storing a computer program that, when executed by a processor, implements the steps of the underground pipeline data desensitization method described in the first aspect.

[0056] This invention acquires underground pipeline control points and uses their coordinates as input data to train a neural network transformation model. It then acquires the data to be anonymized and uses the neural network transformation model to perform anonymization processing, obtaining intermediate data. A perturbation is applied to the intermediate data, and Chebyshev polynomials are used to calculate the model coefficients of the underground pipeline data anonymization model based on the intermediate data and the perturbation. The intermediate data is then input into the underground pipeline data anonymization model to obtain the final anonymized data. This invention employs two different anonymization models to perform multi-level transformation and anonymization of the original underground pipeline data, making it more resistant to attacks and more complex. Simultaneously, an anonymization index is used to control the results after anonymization, improving data confidentiality while ensuring data availability.

[0057] 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 only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort. Attached Figure Description

[0058] Figure 1A flowchart illustrating the steps of an underground pipeline data anonymization method provided in an exemplary embodiment;

[0059] Figure 2 This is a schematic diagram of a module for an underground pipeline data desensitization device provided in an exemplary embodiment;

[0060] Figure 3 This is a schematic diagram of an electronic device provided in one exemplary embodiment. Detailed Implementation

[0061] To make the objectives, technical solutions, and advantages of this application clearer, the embodiments of this application will be described in further detail below with reference to the accompanying drawings.

[0062] It should be understood that the described embodiments are merely some, not all, of the embodiments of this application. All other embodiments obtained by those skilled in the art based on the embodiments of this application without creative effort are within the scope of protection of the embodiments of this application.

[0063] The terminology used in the embodiments of this application is for the purpose of describing particular embodiments only and is not intended to limit the embodiments of this application. The singular forms “a,” “the,” and “the” used in the embodiments of this application and the appended claims are also intended to include the plural forms unless the context clearly indicates otherwise. It should also be understood that the term “and / or” as used herein refers to and includes any or all possible combinations of one or more of the associated listed items.

[0064] In the following description, when referring to the accompanying drawings, unless otherwise indicated, the same numbers in different drawings represent the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with this application. Rather, they are merely examples of apparatuses and methods consistent with some aspects of this application as detailed in the appended claims. In the description of this application, it should be understood that the terms "first," "second," "third," etc., are used only to distinguish similar objects and are not necessarily used to describe a specific order or sequence, nor should they be construed as indicating or implying relative importance. Those skilled in the art can understand the specific meaning of the above terms in this application according to the specific circumstances.

[0065] Furthermore, in the description of this application, unless otherwise stated, "multiple" means two or more. "And / or" describes the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A alone, A and B simultaneously, or B alone. The character " / " generally indicates that the preceding and following related objects have an "or" relationship.

[0066] like Figure 1 As shown in the figure, this application provides a method for desensitizing underground pipeline data, including the following steps:

[0067] S201: Obtain underground pipeline control points, and establish a neural network transformation model based on the coordinates of the underground pipeline control points as input data for training.

[0068] In a preferred embodiment, to ensure that the error during the desensitization process is within a controllable range, a desensitization index needs to be set for the neural network transformation model to limit the desensitization error. Based on the coordinates of the underground pipeline control points, neural network sample points are obtained; based on the coordinates of the neural network sample points, transformed neural network sample points are obtained through the neural network transformation model; then, based on the transformed neural network sample points, the transformation error is calculated. If the difference between the transformation error and the preset transformation index exceeds a preset second threshold, the expansion constant in the neural network transformation model is iteratively processed; when the difference between the transformation error and the preset transformation index is less than the preset second threshold, the iteration of the expansion constant is stopped, and the neural network transformation model is updated according to the current weights. In this embodiment, the desensitization index of the neural network transformation model is set to 1:σ1. When the output of the neural network transformation model satisfies this desensitization index, the adjustment of the neural network transformation model parameters is stopped.

[0069] Specifically, control points for underground pipelines are acquired, and pairs of control points with the same name are selected to provide displacement for the desensitized model. Preferably, a larger disturbance is applied when sensitive pipeline points are used as control points, while smaller disturbances are applied to other control points. Simultaneously, the uniformity of control point selection must be ensured to avoid strong locality in the desensitized data. After the control point pair selection operation is completed, the source control points (fX) for desensitization are obtained. i ,fY i ) and target control point (tX) i ,tY i ), where i = 1, ..., n. The source control point and the target control point do not coincide, and the disturbance of the control point is (ΔX i ,ΔY i ), where i = 1, ..., n.

[0070] The neural network transformation model has a three-layer structure. The first layer is the input two-dimensional coordinate data, with a dimension of 2, representing the X and Y coordinates of the control points, specifically:

[0071]

[0072] The second layer is the hidden layer, with the same number of hidden nodes as the input layer, which is n. The input control points are the centers of the basis functions in the hidden layer. The expansion constant of each basis function is θ, which is determined by the standard deviation after randomly selecting the centers. The larger the value of θ, the smoother the fitting result.

[0073] The minimum bounding rectangle T of the input spatial data is determined, and the maximum distance between any two points in T is:

[0074]

[0075] To make the fitting results as smooth as possible, the expansion constant is chosen as:

[0076]

[0077] The third layer is the output layer, with a dimension of 2, representing the desensitization perturbation of the control point in the X and Y directions, respectively:

[0078]

[0079] By training the neural network transformation model, a weight matrix W with dimension 2 is obtained, which corresponds to the training parameters of the control points in the X and Y directions, respectively, where:

[0080]

[0081] To meet the desensitization index σ1, the above model needs to be checked and the standard error calculated. If the standard error of the sample points after the neural network transformation model does not meet the desensitization index, the expansion constant θ needs to be scaled and the neural network model retrained until the desensitization index requirement is met.

[0082] Specifically, using the smallest bounding rectangle T as the global sample, setting an interval value (interval), and selecting from the smallest bounding rectangle... These points are used as sample points for calculating the mean square error (sX). i ,sY i Substituting the weight W into the following formula:

[0083]

[0084] Obtain the transformed sample points (stX) i ,stY i ).

[0085] The standard error d of the sample points sample The calculation formula is as follows:

[0086]

[0087] If d sampleIf the difference between the target value and the desensitization index σ1 is greater than the limit, it indicates that the expected goal has not been achieved and iteration is required. sample If σ < 1, then the expansion constant θ needs to be increased; otherwise, θ needs to be decreased. The formula for the scaling factor of θ each time is:

[0088]

[0089] After obtaining the new θ, it is fed into the neural network transformation model, the neural network transformation model is retrained, and the weights W are calculated. This process is repeated multiple times until d is reached. sample The difference from σ1 is less than the limit.

[0090] S202: Obtain the data to be de-identified, and perform de-identification processing on the data to be de-identified through the neural network transformation model to obtain intermediate data.

[0091] In a preferred embodiment, a topology-preserving algorithm is also used to ensure that the transformed original pipeline control points maintain the topological relationship of the original pipeline to the greatest extent possible after the nonlinear transformation. Interpolation point coordinates are obtained based on adjacent underground pipeline control points; then, the transformed adjacent underground pipeline control points and the interpolation point coordinates are obtained through the neural network transformation model; a preset point insertion threshold is set; if the distance between the transformed interpolation point coordinates and the straight line containing the transformed adjacent underground pipeline control points exceeds the point insertion threshold, the interpolation point is added to the underground pipeline control points.

[0092] Specifically, the input data for the neural network transformation model is line features, and the coordinate string is (x... i ,y i The threshold for point insertion is φ, and the interval distance is d.

[0093] For any two adjacent points A(x) on the input features i ,y i ) and B(x i+1 ,y i+1 ), calculate its transformation quantity (Δx) i ,Δy i ) and (Δx i+1 ,Δy i+1 ), and obtain the transformed coordinate values ​​A′(x) of A and B from the transformation quantities. i +Δx i ,y i +Δy i ) and B′(x i+1 +Δx i+1 ,y i+1 +Δy i+1 ).

[0094] From the coordinates of A′ and B′, we obtain the equation of the straight line A′B′: ax + by + c = 0, and the coefficients a, b, and c. Calculate the coordinates of the interpolation point C on AB, with the interpolation distance being d. Calculate the desensitization transformation amount (Δx) from the coordinates of C. C ,Δy C ), and obtain C′(x) from the change. C +Δx C ,y C +Δy C Calculate the distance d from line C′ to line A′B′. C .

[0095] If d C If φ > 0, it is assumed that the nonlinear transformation has caused a change in the topology. To maintain the topological relationship, C′ is inserted into the desensitized elements to form A′C′B′. If d C If the value is less than φ, then no action is taken. Find the next interpolation point on AB and repeat the above steps iterating until AB has been traversed.

[0096] S203: Apply a perturbation to the intermediate data, and calculate the model coefficients of the underground pipeline data desensitization model using Chebyshev polynomials based on the intermediate data and the perturbation.

[0097] In a preferred embodiment, control points are obtained based on the selection of intermediate data, and the coordinates of the control points are normalized; a perturbation is applied to the control points to obtain de-identified control points; the coefficients of the Chebyshev polynomial matrix are calculated using the least squares method based on the coordinates of the control points and the coordinates of the de-identified control points; and a de-identification model for underground pipeline data is established based on the Chebyshev polynomial matrix coefficients and the Chebyshev expression of the coordinates of the de-identified control points.

[0098] In a preferred embodiment, to ensure that the data error after desensitization is within a controllable range, a desensitization index is also preset. When the data error after desensitization exceeds the preset desensitization index, the coefficients of the underground pipeline data desensitization model are adjusted by changing the disturbance value. The control point coordinates are used as input data for the underground pipeline data desensitization model, and the disturbance value of the control points is used as output data to obtain the model coefficients to be adjusted. The underground pipeline data desensitization model is updated based on these coefficients. Sample points are obtained based on the intermediate data, and desensitized sample data is obtained using the updated underground pipeline data desensitization model. The desensitization error is calculated based on the desensitization sample data. If the difference between the desensitization error and the preset desensitization index exceeds a preset first threshold, the disturbance value is iteratively processed. When the difference between the desensitization error and the preset desensitization index is less than the preset first threshold, the iteration of the disturbance value is stopped, and the current coefficients of the underground pipeline data desensitization model are determined as the coefficients of the underground pipeline data desensitization model used to calculate the final desensitized data. In this embodiment of the application, the desensitization index for data desensitized by Chebyshev polynomial is 2:σ².

[0099] Specifically, select num control points Cpoint(sX) based on the intermediate data. i ,sY i The intermediate data is used as a global sample. A grid of uniform size is generated within the data range, and the nodes of each grid are used as num sample points (sX) to calculate the desensitization error. i ,sY i Find the smallest bounding rectangle of this intermediate data, denoted as sX. i ∈[a,b],sY i ∈[c,d], used for coordinate normalization calculation. Then, a disturbance (ΔX) is applied to the control points. i ,ΔY i ), to obtain the desensitized control point (tX) i ,tY i This forms a pair of control points before and after desensitization, num. The disturbance value should be set approximately equal to the desensitization index σ2 to reduce the number of iterations for desensitization error control.

[0100] Since the Chebyshev polynomial is defined as a sequence on [-1, 1], establishing a desensitization model based on the Chebyshev polynomial requires normalizing the coordinates of the control points to [-1, 1]. The normalized control point coordinates (sX) ig ,sY ig )for:

[0101]

[0102] Disturbance (ΔX)i ,ΔY i According to Chebyshev's form, the Chebyshev desensitization model is as follows:

[0103]

[0104] Among them, CX ij and CY ij ΔX i with ΔY i Chebyshev polynomial coefficients, T i (sX ig ) and T j (sY ig ) are the i-th and j-th Chebyshev polynomials with the x and y coordinates as independent variables after normalization, respectively, and m and n are the cutoff orders of the Chebyshev polynomials.

[0105] The coordinates of the desensitized control points (tX) i ,tY i This can be expressed as the following formula:

[0106]

[0107] The above equations can be expressed as the following system of linear equations:

[0108]

[0109] The above formula can be simplified to b = A·C, where b is the disturbance matrix of the control points, A is the Chebyshev polynomial matrix, and C is the coefficient matrix. Based on the coordinates of k pairs of control points before and after desensitization, the Chebyshev polynomial coefficient matrix C is obtained using the least squares principle:

[0110] C = (A T A) -1 A T b

[0111] Substituting the coefficient matrix C into the desensitized control point coordinate formula, a preliminary Chebyshev desensitization model is established, which is also a preliminary desensitization model for underground pipeline data.

[0112] To ensure that the accuracy of the desensitized vector data meets the desensitization index requirements, it is necessary to iteratively adjust the control point disturbance amount to obtain the normalized control point coordinates (sX). ig ,sY ig As input data for the desensitization model:

[0113]

[0114] Substituting the input data into the Chebyshev polynomial matrix and using the disturbance at the control points as the output data, we obtain:

[0115]

[0116] The formula for calculating the model coefficient matrix is ​​as follows:

[0117]

[0118]

[0119] Based on the obtained coefficients, a data anonymization model for underground pipelines is established, and its formula is as follows:

[0120]

[0121] Samplepoint(sX) of num sample points i ,sY i After coordinate normalization, the data is substituted into the desensitization model formula for underground pipeline data to calculate the desensitized sample point (tX). i ,tY i And calculate the desensitization error:

[0122]

[0123] The desensitization error is compared with the desensitization index σ2. If |RMSE-σ2| is greater than a certain threshold, the disturbance of the control point needs to be iteratively adjusted. The offset of the control point is scaled as follows:

[0124]

[0125] If RMSE < σ², increase the disturbance (ΔX). i ,ΔY i If RMSE > σ², reduce the disturbance (ΔX). i ,ΔY i The iteration continues until |RMSE-σ2| is less than a specific threshold, at which point the iteration stops, and the model parameters CX and CY are determined and output.

[0126] S204: Input the intermediate data into the underground pipeline data desensitization model to obtain the final desensitized data.

[0127] In a preferred embodiment, after obtaining the final de-identified data, a key file needs to be stored, which includes the control point, the disturbance amount, and the de-identified sample data.

[0128] In a preferred embodiment, in addition to the spatial desensitization process for underground pipeline data described in the above embodiments, the process also includes attribute desensitization and content desensitization of the underground pipeline data. Specifically, attribute desensitization of underground pipelines mainly focuses on removing sensitive information such as plane coordinates, elevation, pipe diameter, pressure, and ancillary structures. Content desensitization of underground pipelines mainly focuses on completely removing underground pipeline data involving important departments.

[0129] This application embodiment acquires underground pipeline control points, uses the coordinates of these control points as input data for training, and establishes a neural network transformation model. It then acquires the data to be de-identified and uses the neural network transformation model to perform de-identification processing on the data, obtaining intermediate data. A perturbation is applied to the intermediate data, and Chebyshev polynomials are used to calculate the model coefficients of the underground pipeline data de-identification model based on the intermediate data and the perturbation. The intermediate data is then input into the underground pipeline data de-identification model to obtain the final de-identified data. This invention employs two different de-identification models to perform multi-level transformation and de-identification of the original underground pipeline data, making it more resistant to attacks and more complex. Simultaneously, de-identification indices are used to control the de-identified results, improving data confidentiality while ensuring data availability.

[0130] This application also provides an underground pipeline data desensitization device 300, such as... Figure 2 As shown, it includes:

[0131] The neural network transformation model establishment module 301 is used to obtain underground pipeline control points and establish a neural network transformation model based on the coordinates of the underground pipeline control points as input data.

[0132] The intermediate data acquisition module 302 is used to acquire data to be desensitized and to perform desensitization processing on the data to be desensitized through the neural network transformation model to acquire intermediate data.

[0133] The model coefficient acquisition module 303 is used to apply a perturbation amount to the intermediate data, and calculate the model coefficients of the underground pipeline data desensitization model based on the intermediate data and the perturbation amount of the intermediate data using Chebyshev polynomials.

[0134] The desensitized data acquisition module 304 is used to input the intermediate data into the underground pipeline data desensitization model to obtain the final desensitized data.

[0135] In one exemplary example, the model coefficient acquisition module 303 further includes:

[0136] The normalization processing unit is used to obtain control points based on the selection of intermediate data and to perform coordinate normalization processing on the control points;

[0137] The desensitized control point acquisition unit is used to apply a disturbance to the control point and acquire the desensitized control point;

[0138] The matrix coefficient acquisition unit is used to calculate the Chebyshev polynomial matrix coefficients using the least squares method based on the coordinates of the control points and the coordinates of the desensitized control points.

[0139] The data anonymization model establishment unit is used to establish an underground pipeline data anonymization model based on the Chebyshev polynomial matrix coefficients and the Chebyshev expression of the anonymized control point coordinates.

[0140] It should be noted that the underground pipeline data desensitization device and the underground pipeline data desensitization method provided in the above embodiments belong to the same concept. For details of their specific implementation process, please refer to the method embodiments, which will not be repeated here.

[0141] like Figure 3 As shown, Figure 3 This is a structural block diagram of an electronic device according to an exemplary embodiment of this application.

[0142] The electronic device includes a processor 910 and a memory 920. The main control chip may contain one or more processors 910. Figure 3 Taking a processor 910 as an example, the number of memory 920s in this main control chip can be one or more. Figure 3 Take a memory chip 920 as an example.

[0143] The memory 920, as a computer-readable storage medium, can be used to store software programs, computer-executable programs, and modules, such as the underground pipeline data desensitization method program described in any embodiment of this application, and the corresponding program instructions / modules for the underground pipeline data desensitization method described in any embodiment of this application. The memory 920 may mainly include a program storage area and a data storage area. The program storage area may store the operating system and at least one application program required for a function; the data storage area may store data created based on the use of the device, etc. Furthermore, the memory 920 may include high-speed random access memory and may also include non-volatile memory, such as at least one disk storage device, flash memory device, or other non-volatile solid-state storage device. In some instances, the memory 920 may further include memory remotely located relative to the processor 910, and these remote memories can be connected to the device via a network. Examples of such networks include, but are not limited to, the Internet, corporate intranets, local area networks, mobile communication networks, and combinations thereof.

[0144] The processor 910 executes various functional applications and data processing of the device by running software programs, instructions and modules stored in the memory 920, thereby implementing a method for desensitizing underground pipeline data as described in any of the above embodiments.

[0145] This application also provides a computer-readable storage medium storing a computer program thereon, which, when executed by a processor, implements a method for desensitizing underground pipeline data as described in any of the above embodiments.

[0146] This invention can take the form of a computer program product implemented on one or more storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing program code. Computer-readable storage media include permanent and non-permanent, removable and non-removable media, and information storage can be implemented by any method or technology. Information can be computer-readable instructions, data structures, program modules, or other data. Examples of computer storage media include, but are not limited to: phase-change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, CD-ROM, digital versatile optical disc (DVD) or other optical storage, magnetic tape, magnetic magnetic disk storage or other magnetic storage devices, or any other non-transfer medium that can be used to store information accessible by a computing device.

[0147] It should be understood that the embodiments of this application are not limited to the precise structures described above and shown in the accompanying drawings, and various modifications and changes can be made without departing from their scope. The scope of the embodiments of this application is limited only by the appended claims.

[0148] The above-described embodiments are merely illustrative of several implementation methods of the embodiments of this application, and their descriptions are relatively specific and detailed, but they should not be construed as limiting the scope of the invention patent. It should be noted that those skilled in the art can make several modifications and improvements without departing from the concept of the embodiments of this application, and these all fall within the protection scope of the embodiments of this application.

Claims

1. A method for desensitizing underground pipeline data, characterized in that, Includes the following steps: Obtain underground pipeline control points, and establish a neural network transformation model based on the coordinates of the underground pipeline control points as input data for training behavior; Obtain the data to be de-identified, and perform de-identification processing on the data to be de-identified through the neural network transformation model to obtain intermediate data; A perturbation amount is applied to the intermediate data, and the model coefficients of the underground pipeline data desensitization model are obtained by Chebyshev polynomial calculation based on the intermediate data and the perturbation amount. The intermediate data is input into the underground pipeline data desensitization model to obtain the final desensitized data; The process, after obtaining the intermediate data, also includes: Control points are obtained based on the selection of intermediate data. The control points are then subjected to coordinate normalization. Apply a disturbance to the control point Obtain control points after desensitization ; Based on the coordinates of the control points and the coordinates of the desensitized control points, the coefficients of the Chebyshev polynomial matrix are calculated using the least squares method. Based on the Chebyshev polynomial matrix coefficients and the Chebyshev expression for the desensitized control point coordinates, a data desensitization model for underground pipelines is established. The Chebyshev expression for the desensitized control point coordinates is: in, and They are respectively and Chebyshev polynomial coefficients, and After naturalization coordinates and The coordinates of the independent variable are the first... Rank and first The Chebyshev polynomial of order m and n is the cutoff order of the Chebyshev polynomial. The process of obtaining the model coefficients of the desensitized underground pipeline data model further includes: The coordinates of the control points are used as the input data of the underground pipeline data desensitization model, and the disturbance of the control points is used as the output data of the underground pipeline data desensitization model to obtain the model coefficients to be adjusted. Update the underground pipeline data desensitization model based on the model coefficients to be adjusted; Sample points are obtained based on the intermediate data, and desensitized sample data is obtained through the updated underground pipeline data desensitization model. Based on the desensitized sample data, the desensitization error is calculated. If the difference between the desensitization error and the preset desensitization index exceeds the preset first threshold, the disturbance amount is iteratively processed. When the difference between the desensitization error and the preset desensitization index is less than the preset first threshold, the iteration of the disturbance amount is stopped, and the coefficients of the current underground pipeline data desensitization model are determined as the coefficients of the underground pipeline data desensitization model used to calculate the final desensitized data.

2. The method for desensitizing underground pipeline data according to claim 1, characterized in that, The establishment of the neural network transformation model also includes: Based on the coordinates of the underground pipeline control points, neural network sample points are obtained. ; Based on the coordinates of the neural network sample points, the transformed neural network sample points are obtained through the neural network transformation model. ; Based on the transformed neural network sample points The transformation error is calculated. If the difference between the transformation error and the preset transformation index exceeds a preset second threshold, the expansion constant in the neural network transformation model is iteratively processed. The formula for calculating the transformation error is: The iterative formula for the extended constant is: in, The transformation error is... For the aforementioned extension constant, The preset transformation index is defined as follows: snum is the number of sample points selected in the minimum bounding rectangle that serves as the global sample, and N is the number of control points. When the transformation error With preset transformation index When the difference is less than a preset second threshold, the expansion constant is stopped. The process iterates and updates the neural network transformation model based on the weights of the current neural network transformation model.

3. The method for desensitizing underground pipeline data according to claim 1, characterized in that, Before obtaining the intermediate data, the process also includes: Based on the adjacent underground pipeline control points, obtain the coordinates of the interpolation point; The neural network transformation model is used to obtain the transformed coordinates of the adjacent underground pipeline control points and the interpolation points. A preset point insertion threshold is set. If the distance between the transformed interpolated point coordinates and the straight line where the adjacent underground pipeline control point is located exceeds the point insertion threshold, then the interpolated point is added to the underground pipeline control point.

4. The method for desensitizing underground pipeline data according to claim 1, characterized in that, After obtaining the final de-identified data, the process also includes: A key file is stored, which includes the control point, the disturbance amount, and the desensitized sample data.

5. The method for desensitizing underground pipeline data according to claim 1, characterized in that, The iterative processing of the disturbance amount specifically includes: The formula for calculating the desensitization error is as follows: Among them, the For the desensitization error, the The x and y coordinates of the sample points are given. ... The disturbance amount The iterative formula is: Among them, the The preset desensitization index; when At that time, increase the disturbance amount ,if Reduce disturbance amount .

6. A data anonymization device for underground pipelines, characterized in that, include: The neural network transformation model establishment module is used to obtain underground pipeline control points and establish a neural network transformation model based on the coordinates of the underground pipeline control points as input data. The intermediate data acquisition module is used to acquire the data to be desensitized, and to desensitize the data to be desensitized through the neural network transformation model to acquire intermediate data; The model coefficient acquisition module is used to apply a perturbation amount to the intermediate data, and to calculate the model coefficients of the underground pipeline data desensitization model based on the intermediate data and the perturbation amount using Chebyshev polynomials. The desensitized data acquisition module is used to input the intermediate data into the underground pipeline data desensitization model to obtain the final desensitized data; Wherein, after the intermediate data acquisition module acquires the intermediate data, the model coefficient acquisition module is further used for: Control points are obtained based on the selection of intermediate data. The control points are then subjected to coordinate normalization. Apply a disturbance to the control point Obtain control points after desensitization ; Based on the coordinates of the control points and the coordinates of the desensitized control points, the coefficients of the Chebyshev polynomial matrix are calculated using the least squares method. Based on the Chebyshev polynomial matrix coefficients and the Chebyshev expression for the desensitized control point coordinates, a data desensitization model for underground pipelines is established. The Chebyshev expression for the desensitized control point coordinates is: in, and They are respectively and Chebyshev polynomial coefficients, and After naturalization coordinates and The coordinates of the independent variable are the first... Rank and first A Chebyshev polynomial of order m and n, where m and n are the cutoff orders of the Chebyshev polynomial; and, The coordinates of the control points are used as the input data of the underground pipeline data desensitization model, and the disturbance of the control points is used as the output data of the underground pipeline data desensitization model to obtain the model coefficients to be adjusted. Update the underground pipeline data desensitization model based on the model coefficients to be adjusted; Sample points are obtained based on the intermediate data, and desensitized sample data is obtained through the updated underground pipeline data desensitization model. Based on the desensitized sample data, the desensitization error is calculated. If the difference between the desensitization error and the preset desensitization index exceeds the preset first threshold, the disturbance amount is iteratively processed. When the difference between the desensitization error and the preset desensitization index is less than the preset first threshold, the iteration of the disturbance amount is stopped, and the coefficients of the current underground pipeline data desensitization model are determined as the coefficients of the underground pipeline data desensitization model used to calculate the final desensitized data.

7. An electronic device, characterized in that, include: At least one memory and at least one processor; The memory is used to store one or more programs; When the one or more programs are executed by the at least one processor, the at least one processor implements the steps of the underground pipeline data desensitization method as described in any one of claims 1 to 5.

8. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by the processor, it implements the steps of the underground pipeline data desensitization method as described in any one of claims 1 to 5.

Images