Positioning method, system and device based on high-precision beidou differential positioning

Abstract

This application provides a positioning method, system, and device based on BeiDou high-precision differential positioning. The method includes: receiving observation data from a reference station network at a central end; obtaining a state vector based on the observation data, the state vector including the reference station coordinate increment, tropospheric grid value, and ionospheric grid value; generating correction packets at regular grid points at preset intervals; transmitting the correction packets to the device coverage area through a one-way broadcast channel to reduce the uplink exposure risk of device-side location information; receiving the correction packets and verifying the signature and timeliness of the power grid inspection device, performing interpolation operations on four neighboring grid points to obtain the covariance after correction and propagation at the device, and constructing an observation weight matrix; the power grid inspection device performing cycle slip detection and ionospheric-free combination processing on local observations; performing a chi-square test on the solution residuals for integrity monitoring, and if the ambiguity is successfully fixed and passes the integrity monitoring, then fixing and resolving based on the integer ambiguity to obtain the positioning result.

Description

Technical Field

[0001] This application relates to the field of BeiDou satellite positioning and correction technology, and in particular to a positioning method, system and equipment based on BeiDou high-precision differential positioning. Background Technology

[0002] With the continuous expansion of the power grid and the widespread application of technologies such as ultra-high voltage power transmission, smart substations, and drone inspections, the demand for high-precision positioning services in power systems is increasing in business scenarios such as equipment inspection, personnel safety management, emergency dispatch, and asset management. Satellite navigation and positioning technology, especially the high-precision positioning capabilities based on the BeiDou system, is gradually becoming an important fundamental supporting technology for the power industry. By constructing a reference station network and providing differential correction information to terminal equipment, positioning accuracy and stability can be effectively improved, enabling power grid inspection equipment to obtain reliable location services even in complex environments, thereby enhancing the intelligence and automation level of power grid operation and maintenance.

[0003] In related technologies, power grid positioning systems typically calculate differential correction information through a network of reference stations and transmit the correction data to terminal devices to assist in positioning calculations. However, existing technologies still have certain shortcomings in terms of large-scale regional service, data security protection, and system stability. For example, during positioning services, terminal devices often need to frequently request location information from upstream sources, which can easily lead to data security risks. Furthermore, the efficiency of transmitting and applying differential correction information over large areas still has room for improvement. In addition, the continuity and reliability of terminal positioning results are easily affected when satellite signals are blocked or communication links are unstable. Summary of the Invention

[0004] In view of this, it is necessary to provide a positioning method, system and equipment based on BeiDou high-precision differential positioning, which can at least overcome one of the above defects.

[0005] In a first aspect, embodiments of this application provide a positioning method based on BeiDou high-precision differential positioning, applied to a BeiDou positioning network composed of power grid inspection equipment and a central terminal, the method comprising:

[0006] The central terminal receives observation data from the reference station network;

[0007] A state vector is obtained based on the observation data, and the state vector includes the base station coordinate increment, tropospheric grid value, and ionospheric grid value.

[0008] Correction packages are generated at preset intervals for regular grid points, and the correction packages include differential corrections and covariance descriptions for the grid points;

[0009] The correction package is sent to the device coverage area through a one-way broadcast channel to reduce the risk of uplink exposure of device-side location information;

[0010] The power grid inspection equipment receives the correction packet and performs interpolation operations on the four neighboring grid points according to the correction packet to obtain the correction number at the equipment and the covariance after propagation;

[0011] Construct an observation weight matrix based on the corrections at the device and the propagated covariance.

[0012] The power grid inspection equipment performs cycle slip detection and ionosphere-free combination processing on local observations, constructs linearized observation equations to obtain floating-point solutions, and applies the LAMBDA method to search for integer ambiguities. A preset ratio threshold is used as the criterion for successful ambiguity fixation.

[0013] Perform a chi-square test on the solution residuals to monitor integrity. If the ambiguity is successfully fixed and passes the integrity monitoring, then fix and resolve based on the integer ambiguity to obtain the positioning result.

[0014] In one embodiment, obtaining the state vector based on the observation data includes:

[0015] Adaptive filtering algorithms or weighted least squares methods are applied to perform quality control and systematic error removal on the raw observations of the reference station network in order to estimate the state vector;

[0016] When generating the correction package, the calculated correction covariance description and grid coordinate information are encapsulated, and the data digest is signed using an asymmetric encryption algorithm.

[0017] In one embodiment, interpolation is performed on the four neighboring grid points to obtain the device-corrected and propagated covariance, and an observation weight matrix is ​​constructed, including:

[0018] Based on the approximate current location of the power grid inspection equipment, a preset number of grid points are selected in the vicinity, and bilinear interpolation or spatial correlation interpolation algorithm is executed to extract the delay correction term of the location of the power grid inspection equipment.

[0019] The covariance description of the grid points is transmitted to the power grid inspection equipment by applying the error propagation law, and the weight matrix that can dynamically reflect the different satellite observation quality is constructed by combining the local preset random model and the residual error of the interpolation model.

[0020] In one embodiment, the step of applying the LAMBDA method to search for integer ambiguities, using a preset ratio threshold as the criterion for successful ambiguity fixation, includes:

[0021] Construct a combined observation system of wide and narrow alleyways;

[0022] The original ambiguity covariance matrix is ​​decorrelated by applying integer Gaussian transform, which maps the search space to a transform space, and the optimal integer candidate solution is determined in the transform space.

[0023] The conditions for determining ambiguity fixation failure include: the ratio of the optimal candidate solution to the second-best candidate solution does not reach the preset safety threshold, or the chi-square test statistic of the calculated residual components exceeds the limit corresponding to the significance level; when the above failure conditions, communication link interruption, or grid data timeliness failure is detected, the positioning mode is triggered to switch autonomously from carrier phase differential mode to degraded fusion mode.

[0024] In one embodiment, the method further includes:

[0025] If the determination of ambiguity fixation failure, integrity monitoring abnormality, or communication interruption occurs, a downgrade fusion strategy will be triggered.

[0026] The degradation fusion strategy includes outputting positioning results with accuracy confidence labels by fusing inertial measurement unit pre-integration, visual odometry, and historical grid trajectories.

[0027] In one embodiment, the method further includes:

[0028] The inertial measurement unit built into the power grid inspection equipment is invoked to perform high-frequency motion pre-integration, and relative pose estimation is performed by combining the displacement of feature points captured by the visual sensor.

[0029] The fixed grid trajectory within the historical period is injected into the fusion filter as a priori motion constraint to compensate for positioning drift in signal occlusion environments.

[0030] The output includes a continuous stream of location information, including the current solution status, estimation accuracy, availability classification labels, and anomaly audit logs.

[0031] In one embodiment, the method further includes:

[0032] The central end dynamically adjusts the grid coverage density of the local area based on the regional ionospheric activity index and topographic relief characteristics, and uses a polynomial fitting model to extract the large-scale atmospheric delay gradient coefficient and encapsulate it in the correction package.

[0033] When performing interpolation, the power grid inspection equipment performs nonlinear error compensation on the linear interpolation results based on the coefficients of the polynomial fitting model, so as to reduce the non-modeling impact of atmospheric delay residue on carrier phase observations.

[0034] In one embodiment, the method further includes:

[0035] When the power grid inspection equipment identifies that the quality of the one-way broadcast signal is lower than the preset limit or is in a specific authorized task area, it requests information from the uplink location of the central terminal through a secure encrypted channel.

[0036] In response to the request, the central terminal calculates and generates virtual reference station observation data for the corresponding location, and feeds back customized differential correction parameters through a two-way communication channel;

[0037] The central terminal is also used to write the complete interaction process, including identity tokens, request parameters, and distribution records, into the audit log.

[0038] Secondly, one embodiment of this application provides a positioning system based on BeiDou high-precision differential positioning, the system comprising:

[0039] The central terminal is used to receive observation data from the reference station network; obtain a state vector based on the observation data, the state vector including the reference station coordinate increment, tropospheric grid value, and ionospheric grid value; generate correction packets at regular grid points at preset intervals, the correction packets including the differential correction number and covariance description of the grid points; and send the correction packets to the equipment coverage area through a one-way broadcast channel to reduce the uplink exposure risk of equipment-side location information.

[0040] The power grid inspection equipment is used to receive the correction packet, perform interpolation on the four neighboring grid points to obtain the correction number at the equipment location and the propagated covariance, construct an observation weight matrix based on the correction number at the equipment location and the propagated covariance, perform cycle slip detection and ionospheric-free combination processing on the local observations, construct a linearized observation equation to obtain a floating-point solution, and apply the LAMBDA method to search for integer ambiguities, using a preset ratio threshold as the criterion for successful ambiguity fixation; perform a chi-square test on the solution residuals for integrity monitoring, and if the ambiguity is successfully fixed and passes the integrity monitoring, then fix and resolve based on the integer ambiguity to obtain the positioning result.

[0041] Thirdly, embodiments of this application provide a power grid inspection device, comprising:

[0042] Processor; and

[0043] A memory having computer-readable instructions stored thereon for controlling the processor to execute the positioning method based on BeiDou high-precision differential positioning as described in the first aspect.

[0044] This application provides a positioning method, system, and device based on BeiDou high-precision differential positioning. By uniformly solving the observation data of the reference station network at the central end and generating gridded differential correction information, the correction data is then sent to the power grid inspection equipment through a one-way broadcast method. This allows the terminal to obtain high-precision positioning services without continuously uplinking location information, thereby improving positioning accuracy and real-time performance while reducing the risk of equipment-side location information exposure and enhancing the system's data security and traceability. Attached Figure Description

[0045] Figure 1 This is a schematic flowchart of an ionospheric error correction method based on BeiDou satellite signals provided in an embodiment of this application.

[0046] Figure 2 A schematic diagram of a module for an ionospheric error correction system based on BeiDou satellite signals is provided for one embodiment of this application.

[0047] Figure 3 This is a schematic diagram of a power grid inspection device provided in one embodiment of this application.

[0048] Explanation of main component symbols

[0049] Positioning system based on BeiDou high-precision differential positioning 10 Central end 11 Power grid inspection equipment 20 processor 21 memory 22 Methods and Steps S100-S700 Detailed Implementation

[0050] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of them.

[0051] It should be noted that, in the embodiments of this application, "at least one" refers to one or more, and "more than one" refers to two or more. Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs. The terminology used in the specification of this application is for the purpose of describing particular embodiments only and is not intended to be limiting of this application.

[0052] It should be noted that in the embodiments of this application, the terms "first," "second," etc., are used only for descriptive purposes and should not be construed as indicating or implying relative importance, nor as indicating or implying order. Features specified as "first" or "second" may explicitly or implicitly include one or more of the stated features. In the description of the embodiments of this application, words such as "exemplary" or "for example" are used to indicate examples, illustrations, or explanations. Any embodiment or design scheme described as "exemplary" or "for example" in the embodiments of this application should not be construed as being more preferred or advantageous than other embodiments or design schemes. Specifically, the use of words such as "exemplary" or "for example" is intended to present the relevant concepts in a concrete manner.

[0053] Based on the embodiments described in this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0054] With the continuous expansion of the power grid and the widespread application of technologies such as ultra-high voltage power transmission, smart substations, and drone inspections, the demand for high-precision positioning services in power systems is increasing in business scenarios such as equipment inspection, personnel safety management, emergency dispatch, and asset management. Satellite navigation and positioning technology, especially the high-precision positioning capabilities based on the BeiDou system, is gradually becoming an important fundamental supporting technology for the power industry. By constructing a reference station network and providing differential correction information to terminal equipment, positioning accuracy and stability can be effectively improved, enabling power grid inspection equipment to obtain reliable location services even in complex environments, thereby enhancing the intelligence and automation level of power grid operation and maintenance.

[0055] In related technologies, power grid positioning systems typically calculate differential correction information through a network of reference stations and transmit the correction data to terminal devices to assist in positioning calculations. However, existing technologies still have certain shortcomings in terms of large-scale regional service, data security protection, and system stability. For example, during positioning services, terminal devices often need to frequently request location information, which can easily lead to data security risks. Furthermore, the efficiency of transmitting and applying differential correction information over large areas still has room for improvement. In addition, when satellite signals are blocked or communication links are unstable, the continuity and reliability of terminal positioning results are easily affected. Therefore, how to improve the stability, reliability, and regional coverage of differential positioning services while ensuring data security has become a crucial issue that high-precision positioning technology in the power industry needs to further address.

[0056] In view of this, the positioning method, system, and equipment based on BeiDou high-precision differential positioning provided in this application perform unified calculation on the observation data of the reference station network at the central end and generate gridded differential correction information. Then, the correction data is transmitted to the power grid inspection equipment through a one-way broadcast method, so that the terminal can obtain high-precision positioning services without continuously uploading location information. This improves positioning accuracy and real-time performance while reducing the risk of equipment-side location information exposure, and enhances the data security and traceability of the system. At the same time, by introducing covariance description into the correction data and combining it with terminal-side interpolation calculation and weight construction mechanism, the observation quality can be dynamically modeled, thereby improving the stability and reliability of positioning calculation. In addition, by combining ambiguity fixed judgment and integrity monitoring mechanism, the reliability of positioning results can be effectively guaranteed. In the case of signal blockage or communication anomalies, the positioning mode can be adaptively degraded through multi-source information fusion, ensuring the continuous positioning capability of inspection equipment in complex environments, thereby improving the overall safety, stability, and intelligence level of power grid inspection operations.

[0057] Figure 1 This is a schematic flowchart of a positioning method based on BeiDou high-precision differential positioning provided in an embodiment of this application. Figure 1 The positioning method based on BeiDou high-precision differential positioning shown includes at least the following steps: S100: The central end receives observation data from the reference station network; S200: Obtains the state vector based on the observation data; S300: Generates correction packets at regular grid points at preset intervals; S400: Sends the correction packets to the equipment coverage area through a one-way broadcast channel to reduce the uplink exposure risk of equipment-side location information; S500: The power grid inspection equipment receives the correction packets and performs interpolation operations on the four neighboring grid points to obtain the corrected and propagated protocol at the equipment location. Variance: Construct an observation weight matrix based on the corrections at the equipment and the covariance after propagation; S600: The power grid inspection equipment performs cycle slip detection and ionosphere-free combination processing on local observations, constructs a linearized observation equation to obtain a floating-point solution, and applies the LAMBDA method to search for integer ambiguities, using a preset ratio threshold as the criterion for successful ambiguity fixation; S700: Perform a chi-square test on the solution residuals to monitor integrity. If the ambiguity is successfully fixed and passes the integrity monitoring, then fix and resolve based on the integer ambiguity to obtain the positioning result.

[0058] S100: The central terminal receives observation data from the reference station network.

[0059] In this embodiment of the application, the method includes step S100, in which the central end receives observation data from the reference station network.

[0060] Specifically, the central station receives and stores raw observation sets from each reference station, including pseudorange data. carrier phase Observation timestamp Receiver status and antenna markings, observation frequency markings, and corresponding approximate coordinates of the base station. After reception, the observations undergo preliminary preprocessing: time alignment, receiver / antenna phase center deviation (PCO / PCV) correction, receiver clock bias coarse calibration, removal of missing observations and outliers, and recording of metadata (serial number, source site, reception time) for the received batch.

[0061] Understandably, the S100 processing aims to ensure the temporal consistency and basic quality of intra-network observations—without this preprocessing, subsequent intra-network calculations and gridding corrections would introduce avoidable systematic errors and time offsets; at the same time, it preserves complete metadata to facilitate subsequent auditing and backtracking.

[0062] S200: Obtain the state vector based on the observation data.

[0063] In this embodiment of the application, the method includes step S200, which involves obtaining a state vector based on the observation data. The state vector includes the base station coordinate increment, tropospheric grid value, ionospheric grid value, and satellite clock error.

[0064] Specifically, the central end uses the aforementioned observations as input to establish the network observation equations and estimates the state vector using adaptive filtering (e.g., Kalman filtering) or weighted least squares method.

[0065]

[0066] in Indicates the base station The coordinate correction amount (m). and These represent the tropospheric and ionospheric delay parameter vectors (m or delay unit) at the grid nodes, respectively. This represents the satellite clock error / orbit residual correlation term (m). If Kalman filtering is used, the prediction / update form is as follows:

[0067] predict:

[0068] Covariance prediction:

[0069] renew:

[0070] .

[0071] in, —State transition matrix, —Process noise covariance (reflecting the uncertainty of state evolution). —Observation matrix (composed of satellite geometric derivatives) —Observation noise covariance, —Observation residual vector.

[0072] Understandably, the state vector obtained through the above estimation not only includes the correction of the base station coordinates, but also represents the atmospheric field (troposphere / ionosphere) and satellite clock bias in a grid or parameterized form. This provides a physically consistent and quantifiable input for subsequent grid-based corrections and their use in terminal interpolation. At the same time, the covariance output of the filtering process provides a basis for the correction uncertainty.

[0073] S300: Generates a correction package at the regular grid points according to the preset spacing.

[0074] In this embodiment, step S300 includes generating a correction package at a preset interval for each regular grid point. The correction package includes the differential correction number and covariance description of the grid points. In other embodiments, the correction package also includes a timestamp, a sequence number, and an ECDSA-based digital signature.

[0075] Specifically, the central end operates within the coverage area according to a preset grid spacing. Construct a rule-based grid, for each grid point Output differential correction (Including tropospheric correction) Ionospheric correction Satellite clock / orbit correction (etc.) and the corresponding covariance matrix The corrections for each grid point are calculated along with the covariance, grid point latitude / longitude coordinates, and a generated timestamp. The main payload of the correction package includes serial number (seq), version number (ver), etc. The following is a summary of the load calculation: (e.g., SHA-256), and using the central private key. Generate ECDSA signature: This ultimately forms a deployable correction package. .

[0076] Understandably, the covariance description in the correction packet allows the inspection equipment to quantitatively construct observation weights during interpolation and solution. In other embodiments, timestamps and sequence numbers ensure data freshness and verifiable order; ECDSA-based signature and digest mechanisms guarantee the integrity and non-repudiation of the correction packet during transmission, thereby meeting traceability and auditing requirements.

[0077] S400: The correction package is sent to the equipment coverage area through a one-way broadcast channel to reduce the risk of uplink exposure of equipment location information.

[0078] In this embodiment of the application, the method includes step S400, which involves sending the correction package to the device coverage area through a one-way broadcast channel to reduce the uplink exposure risk of device-side location information.

[0079] Specifically, the central terminal sends the signed correction packets to the coverage area periodically or by version update via a one-way downlink (controlled broadcast, private network downlink, or downlink-only multicast channel). This distribution mechanism does not require the inspection equipment to perform the necessary uplink confirmation; the inspection equipment only passively reads the correction after receiving and verifying the signature. Control plane management (such as subscription permissions and key updates) can be completed through a separate controlled bidirectional channel or during operation and maintenance periods to avoid uplink exposure under normal circumstances.

[0080] Understandably, one-way broadcasting significantly reduces the exposure of the inspection device's uplink location information and the man-in-the-middle attack surface (because the differential correction required for daily positioning does not depend on the inspection device's uplink), while still retaining a controlled two-way channel for authorized requests or maintenance operations, providing an engineering compromise between "availability" and "minimum exposure" from a security and compliance perspective.

[0081] S500: The power grid inspection equipment receives the correction packet, performs interpolation on the four neighboring grid points to obtain the correction at the equipment and the covariance after propagation, and constructs the observation weight matrix based on the correction at the equipment and the covariance after propagation.

[0082] In this embodiment of the application, the method includes step S500, in which the power grid inspection equipment receives the correction packet, performs interpolation operation on the four neighboring grid points to obtain the correction and the covariance after propagation at the equipment, and constructs the observation weight matrix based on the correction number at the equipment and the covariance after propagation.

[0083] Specifically, the inspection equipment is located at a rough current position. Select four neighboring grid points (bottom left, bottom right, top left, top right) covering this location and perform bilinear interpolation:

[0084]

[0085] When the covariance is calculated using the error propagation law, it is approximately:

[0086]

[0087] in This represents the residual covariance of the interpolation model (which can be given based on experience or historical statistics). Then... Construct the observation weight matrix together with the local observation noise model: ,in From Uncertainty projection for different satellite corrections.

[0088] — Bilinear interpolation weights (dimensionless). —No. The correction vector for each grid point —Corresponding grid-point corrected covariance matrix (unit: or (represented by covariance).

[0089] Understandably, by propagating the grid covariance to the inspection equipment and incorporating it into the observation weight matrix, the solution of the inspection equipment can dynamically reflect the uncertainty of the correction and the difference in observation quality, thus providing a more robust estimate in the weighted least squares or filtering process rather than simply using the correction as a deterministic value.

[0090] S600: The power grid inspection equipment performs cycle slip detection and ionosphere-free combination processing on local observations, constructs linearized observation equations to obtain floating-point solutions, and applies the LAMBDA method to search for integer ambiguities. A preset ratio threshold is used as the criterion for successful ambiguity fixation.

[0091] In this embodiment of the application, the method includes step S600, in which the power grid inspection equipment performs cycle slip detection and ionosphere-free combination processing on local observations, constructs a linearized observation equation to obtain a floating-point solution, applies the LAMBDA method to search for integer ambiguity, and uses a preset ratio threshold as the determination condition for successful ambiguity fixation.

[0092] Specifically, the inspection equipment performs cycle slip detection on the raw multi-frequency observations (e.g., based on phase and pseudorange residuals, or Melbourne–Wubbena width-to-narrow difference signal verification); for cycle slip-free observations, it constructs ionospheric-free combinations (LC) or other applicable combinations to reduce first-order ionospheric errors.

[0093]

[0094] Correction after interpolation Correct the observations and establish linearized observation equations: Using weight matrix Floating-point estimation For floating-point ambiguity The LAMBDA method is used for integer search, i.e., minimizing the objective function:

[0095]

[0096] in Let be the covariance matrix of the floating-point ambiguity. To improve search efficiency, first... Perform an integer Gaussian transform (recorrelation reduction), search for several optimal candidate integer solutions in the transformed space, and calculate the cost ratio between the best and second-best candidate solutions (i.e., ratio test). If the cost ratio is greater than a pre-set "ratio threshold" and other consistency measures are met (such as reasonable residuals after projecting the candidate solution into the location domain), then the ambiguity is determined to be successfully fixed.

[0097] in, —Two carrier frequencies (Hz); —Design matrix for location and ambiguity; —Floating-point ambiguity covariance (unit: “Ratio threshold” – a preset safety threshold (dimensionless, used to determine the discriminative power of candidate solutions).

[0098] Understandably, this step, by establishing a statistical confidence judgment between floating-point estimation and integer search, can obtain reliable integer fixation under most normal observation conditions; when the observation quality is insufficient or the correction uncertainty is too high, the ratio test will prevent erroneous fixation, thereby avoiding large deviations caused by forcibly fixing erroneous integers.

[0099] S700: Perform a chi-square test on the solution residuals to monitor integrity. If the ambiguity is successfully fixed and passes the integrity monitoring, then fix and resolve based on the integer ambiguity to obtain the positioning result.

[0100] In this embodiment of the application, the method includes step S700, which involves performing a chi-square test on the solved residuals to monitor integrity. If the ambiguity is successfully fixed and passes the integrity monitoring, the positioning result is obtained by fixing and resolving the ambiguity based on the integer ambiguity.

[0101] Specifically, after obtaining candidate integer solutions and deciding to fix them, the inspection equipment calculates the solution residual vector. (in (For parameter estimation with fixed integer ambiguity), and calculate the chi-square statistic using the following formula:

[0102]

[0103] in To observe the covariance matrix (including the corrected covariance component). Based on the set significance level. With degrees of freedom (Usually the number of observations minus the number of parameters to be estimated), comparison With threshold .when Furthermore, when the ambiguity fixation determination passed, the inspection equipment accepted the fixation result and performed a re-resolution with integer ambiguity to output the final positioning solution. and its covariance The output should also include status flags (fixed / floating / degraded), estimated accuracy metrics, and audit logs (the sequence number of the correction package used, timestamp, signature verification results, etc.).

[0104] in, —Residual vector (m) —Observation covariance matrix ( ), —degrees of freedom —Significance level (e.g., used to determine whether something is abnormal).

[0105] Understandably, the chi-square integrity check serves as the final statistical check after the ambiguity is fixed. It can capture systematic errors caused by observational anomalies or inconsistent corrections, and together with the ambiguity determination, it can decide whether to accept the current solution as a reliable high-precision output. If the check fails, the fixed result is not accepted and a downgrade or retry strategy is triggered to ensure positioning security and reliability.

[0106] In this embodiment, obtaining the state vector based on observation data includes: applying an adaptive filtering algorithm or weighted least squares method to perform quality control and systematic error removal on the raw observations of the reference station network to estimate the state vector. When generating the correction package, the calculated correction covariance description and grid point coordinate information are encapsulated, and the data digest is signed using an asymmetric encryption algorithm.

[0107] Specifically, the central end first processes the raw observations from the base station. Quality control is performed (including time alignment, PCO / PCV correction, pseudorange / phase cross-checking, missing data imputation, and outlier observation removal). Then, the intra-network observation equations are established and solved using adaptive Kalman filtering or weighted least squares. Taking the state equations and observation equations as an example, the state prediction and update forms are as follows:

[0108]

[0109]

[0110] in The state vector to be estimated (including the coordinate increment of the base station) tropospheric grid values ionospheric grid value and satellite clock bias wait); Here is the state transition matrix. For process noise covariance, For the observation matrix, To observe the noise covariance, To estimate the covariance of the state, For Kalman gain, To observe the residual vector. After estimation, the center point is located at each grid point. Output correction vector and the corresponding covariance matrix (Depend on (The corresponding sub-blocks are obtained through incremental variance propagation). When generating the correction package, the correction amount, grid coordinates, generation timestamp, and sequence number of each grid point are encapsulated into a payload according to a predetermined format. Calculation summary (e.g., SHA-256), and using the central asymmetric private key. Sign The final correction package For distribution.

[0111] Understandably, this process ensures that the corrections used for distribution are based on physically consistent estimates after quality control and system error removal, while also encapsulating the estimation uncertainty (covariance) along with time / version information to facilitate the quantification of correction uncertainty by inspection equipment. At the same time, summarizing and signing the load ensures the integrity, verifiability, and auditability of the correction package during transmission and use, thereby meeting compliance and security requirements.

[0112] In this embodiment, interpolation is performed on four neighboring grid points to obtain the corrected and propagated covariance at the device location, and an observation weight matrix is ​​constructed. This includes: selecting a preset number of neighboring grid points based on the approximate current location of the power grid inspection device, and performing bilinear interpolation or spatial correlation interpolation algorithms to extract the delay correction term for the location of the power grid inspection device. The covariance description of the grid points is transferred to the power grid inspection device using the error propagation law, and combined with a locally preset stochastic model and the residual error of the interpolation model, a weight matrix that can dynamically reflect the different satellite observation qualities is constructed.

[0113] Specifically, the inspection equipment is based on its rough location. Determine the four neighboring grid points containing this location. Calculate the correction vector at the location using bilinear interpolation:

[0114]

[0115] Among them, weight The planar coordinate offset of the inspection equipment relative to the grid points is calculated according to the bilinear rule; if spatially correlated interpolation (such as kriging) is used, the spatial covariance function between grid points is utilized during interpolation. Weighted interpolation is performed, and the result is a minimum variance unbiased estimate. After interpolation, the propagation of the corrected covariance is calculated according to the error propagation law:

[0116]

[0117] in For the corrected covariance of the grid points, The interpolation model residual covariance is calculated (this can be estimated by historical statistics or empirical models for inspection equipment). Subsequently, the inspection equipment projects the interpolated covariance onto the observation space to construct the correction for the resulting observation uncertainty. (For example, through the Jacobian matrix) : ), and the covariance of local observation noise The overall observation weight matrix is ​​obtained by superposition. In actual implementation, This indicates the projection relationship of the correction vector onto each satellite observation (dimensionless / unit transformation factor depends on the specific correction type).

[0118] Understandably, by transferring and projecting the grid covariance onto the observation weight matrix, the solution algorithm of the inspection equipment can automatically reduce the weights from uncertain corrections or low-quality grid points in weighted least squares or filtering, thereby avoiding unreasonable "overconfident" solutions caused by errors or high uncertainty corrections; using spatial correlation interpolation (such as kriging) can also improve interpolation accuracy in complex terrain or areas with severe ionospheric fluctuations.

[0119] In this embodiment, the LAMBDA method is applied to search for integer ambiguities, and a preset ratio threshold is used as the criterion for successful ambiguity fixing. This includes constructing a combination of wide-lane and narrow-lane observations. An integer Gaussian transform is applied to reduce the correlation of the original ambiguity covariance matrix, mapping the search space to a transform space, and determining the optimal integer candidate solution in the transform space. Conditions for determining ambiguity fixing failure include: the ratio of the optimal candidate solution to the second-best candidate solution does not reach a preset safety threshold, or the chi-square test statistic of the calculated residual components exceeds the limit corresponding to the significance level. When the above failure conditions, communication link interruption, or grid data timeliness failure is detected, an autonomous switching of the positioning mode from carrier phase differential to degraded fusion mode is triggered.

[0120] Specifically, the inspection equipment can first construct combinations of wide-lane and narrow-lane quantities to improve the solvability of ambiguity, for example, expressed in linear combination form as follows: (coefficient (Selecting an option to increase the equivalent wavelength to improve integer resolvability), and then estimating the floating-point ambiguity. and its covariance matrix Applying integer Gaussian transform (decorrelation transform) to obtain the reduced covariance (in (where Gaussian transformation matrix is ​​an integer), and a search is performed in the transformation space to minimize the objective function:

[0121]

[0122] During the verification phase, the cost ratio between the optimal candidate solution and the second-best candidate solution is calculated. If the ratio is greater than a preset threshold, the discrimination is considered sufficient, and the solution is accepted for fixing. Simultaneously, the fixed residual vector can be... Calculate the component chi-square statistics to check for local consistency. If the ratio does not reach the threshold, the chi-square test exceeds the limit, or the inspection equipment detects anomalies such as expired correction packets / link interruption, the ambiguity fixing is considered to have failed, and the inspection equipment automatically switches to the degraded fusion mode: enabling IMU pre-integration, visual odometry and historical grid prior filtering fusion to ensure short-term positioning continuity.

[0123] Understandably, LAMBDA, combining wide and narrow difference signals with integer Gaussian transform, can quickly and reliably obtain integer ambiguity under most good observation conditions; ratio test and chi-square consistency test provide statistical assurance against erroneous fixation; when these statistical tests or communication / correction timeliness conditions are not met, timely switching to a degradation scheme based on other sensors or priors can avoid high-confidence positions of output errors, thereby improving the overall security and robustness of the system.

[0124] In this embodiment, the method further includes: if ambiguity fixing fails, integrity monitoring is abnormal, or communication is interrupted, a degradation fusion strategy is triggered. The degradation fusion strategy includes outputting a positioning result with an accuracy confidence label by fusing inertial measurement unit pre-integration, visual odometry, and historical grid trajectory.

[0125] Specifically, after detecting any triggering condition (such as LAMBDA fixing failure, χ² integrity check failure or correction packet timeout, downlink packet loss / interruption), the inspection equipment enters the degradation fusion process: first, it calls the IMU sensor sequence for pre-integration to obtain the pre-integrated quantity. (representing rotation increment, velocity increment, and position increment respectively), and simultaneously reading the inter-frame feature matching displacement provided by visual odometry. Secondly, trajectory segments in the historical grid that have been determined and marked as "reliable" in recent times are used as prior constraints. The above quantities are fused using a state estimator (e.g., an extended Kalman filter or factor graph minimization) to estimate the state vector. The fusion process weights each observation source according to its covariance: the IMU pre-integration covariance is derived from the pre-integration propagation formula; the visual odometry covariance is measured by the feature matching quality; and the historical grid prior covariance is given by the localization covariance of the prior trajectory. The fusion output also includes an accuracy confidence label (e.g., a level identifier determined based on the trace or principal direction covariance of the estimated covariance) and forms a continuous localization information stream with timestamps and source descriptions.

[0126] Understandably, this downgrade fusion strategy uses short-term, high-frequency inertial constraints and visual relative observations to compensate for the risk of position jumps or drifts caused by carrier phase failures. It also uses historical grid trajectories to provide long-term stability and reference benchmarks, so that even when communication or satellite observation is limited, it can still output continuous and quantifiable confidence positioning results, ensuring the continuity and safety of inspection tasks.

[0127] In this embodiment, the method further includes: invoking the inertial measurement unit built into the power grid inspection equipment to perform high-frequency motion pre-integration, and combining the displacement of feature points captured by the visual sensor to perform relative pose estimation. Fixed grid trajectories from historical time periods are injected as prior motion constraints into the fusion filter to compensate for positioning drift under signal obstruction environments. The output includes a continuous positioning information stream comprising the current solution status, estimation accuracy, availability classification label, and anomaly audit logs.

[0128] Specifically, the IMU pre-integration measures the accelerometer and gyroscope according to the discrete time series. Perform integration and eliminate the effects of measurement bias:

[0129]

[0130]

[0131] in These are the biases for the accelerometer and gyroscope, respectively. For the exponential mapping from Euler vectors to rotation matrices, For reference time to time The rotation matrix. Visual odometry obtains the inter-frame relative pose through feature detection and matching. and matching confidence; historical grid trajectory The prior constraint term is transformed and its covariance is injected into the prior covariance matrix of the fusion unit. The fusion is solved using a Kalman or nonlinear least squares (factor graph) method with priors, and the output is: current position. ,attitude Estimating covariance The audit log consists of availability grading (e.g., "high / medium / low" labels, determined by covariance or estimated uncertainty range), the reason for this downgrade trigger, and the data packet sequence number used.

[0132] Understandably, IMU pre-integration provides short-term constraints to maintain high-frequency dynamic tracking, visual odometry provides observation closure to suppress long-term drift, and historical grid trajectory serves as a reference to help recover absolute scale and position when vision fails or features are scarce. This enables the inspection equipment to provide confident positioning output even in complex electromagnetic or occlusion environments, and to retrospectively review positioning quality and decision-making processes through audit logs.

[0133] In this embodiment, the method further includes: dynamically adjusting the grid coverage density of a local area based on the regional ionospheric activity index and topographic undulation characteristics at the central end, and extracting large-scale atmospheric delay gradient coefficients using a polynomial fitting model and encapsulating them in a correction package. When performing interpolation, the power grid inspection equipment performs nonlinear error compensation on the linear interpolation results based on the polynomial fitting model coefficients to reduce the unmodeled impact of atmospheric delay residues on carrier phase observations.

[0134] Specifically, the central endpoint estimates regional ionospheric / tropospheric activity indices based on historical and real-time observations and generates density scheduling strategies by combining topographic relief (slope, elevation changes, etc.), thereby deploying denser grids or improving correction update rates in active or complex terrain regions; for large-scale atmospheric delay fields, the central endpoint models the delay field in a two-dimensional polynomial form in a local coordinate system:

[0135]

[0136] in This represents the atmospheric delay (m) at the grid plane coordinates. These are the fitting coefficients. The order of the polynomial. Fitting coefficients. The fitting residual statistics are encapsulated together in the correction package so that the inspection equipment can refer to this model for nonlinear compensation during local interpolation: the inspection equipment initially obtains the model through bilinear or spatial correlation interpolation. Then correct it according to the model. ,in This is a polynomial approximation for the interpolation prediction. The correction package also carries the model uncertainty so that the inspection equipment can adjust the interpolation covariance accordingly.

[0137] Understandably, by extracting and distributing polynomial coefficients that characterize large-scale delay trends at the central end, the inspection equipment can utilize local information from grid interpolation and obtain compensation for large-scale nonlinear components locally. This reduces residual errors caused by changes in grid spacing or atmospheric gradients, significantly improving positioning accuracy and stability in areas of drastic ionospheric / tropospheric changes.

[0138] In this embodiment, the method further includes: when the power grid inspection equipment identifies that the quality of the unidirectional broadcast signal is lower than a preset limit or is located in a specific authorized task area, it requests uplink location information from the central terminal through a secure encrypted channel. In response to the request, the central terminal calculates and generates virtual reference station observation data for the corresponding location and feeds back customized differential correction parameters through a bidirectional communication channel. The central terminal is also used to write the complete interaction process, including the identity token, request parameters, and distribution records, into an audit log.

[0139] Specifically, when the inspection equipment detects that the received signal quality indicators (such as signal-to-noise ratio, packet loss rate, or signature verification failure rate) are below a preset threshold, or when performing a specific authorized task (requiring higher precision or operation in a restricted area), it constructs an uplink request message containing an identity token, the current coordination timestamp, the inspection equipment's location information, and the request type. This message is sent to the central end through an established secure encrypted channel (e.g., based on mutually authenticated TLS+ client certificates or a key-based lightweight encrypted channel). After verifying the identity and permissions, the central end generates a virtual reference station (VRS) observation data packet (including virtual observation pseudorange / phase, corresponding covariance, and signature) for the corresponding location based on observations from nearby reference stations or intra-network calculation results, and returns it to the inspection equipment through a controlled bidirectional channel. Simultaneously, the central end writes a complete record of the requested identity token, request parameters, response packet sequence number, signature hash, and response time into the audit log for traceability. After receiving the VRS data, the inspection equipment verifies the signature and performs local calculations using the customized correction, either replacing or overlaying the grid correction.

[0140] Understandably, while ensuring a minimal exposure security strategy that primarily relies on one-way broadcasting in daily operations, this mechanism still provides a controlled two-way on-demand service channel for special situations or signal degradation scenarios. Through strong authentication and audit chain recording, it not only meets the business's need for high-precision customization but also ensures the traceability and non-repudiation of all interactions in terms of compliance, thus balancing security and functionality.

[0141] The following exemplary embodiment describes the complete workflow of the positioning method based on BeiDou high-precision differential positioning provided in this application.

[0142] This method is based on the multi-station observation system and central-end solution deployment of the BeiDou Navigation Satellite System: The central end continuously receives raw observations (pseudorange, carrier phase, observation time, station and antenna metadata, etc.) reported by each reference station in the coverage area. These observations are first time aligned, antenna / receiver correlated, abnormal observations are removed and missing values ​​are marked. Then, the observation equation is established in the network and a unified system state vector is estimated. This state vector includes reference station coordinate correction terms, atmospheric delay fields (troposphere and ionosphere) and satellite clock errors / orbit residuals represented by regular grids, and outputs the estimated uncertainty (covariance description). The central terminal encapsulates the correction amount and its covariance, grid coordinates, generation timestamp and sequence number of each grid point into a correction packet. It calculates a digest for the packet body and signs it with an asymmetric private key (to ensure source and integrity). After writing the signed correction packet into the audit log, it periodically or by version is distributed through a controlled downlink channel in a one-way broadcast manner. This one-way broadcast logic allows the terminal to not need uplink location information under normal circumstances, thereby reducing the risk of uplink exposure. However, under controlled / authorized circumstances, the option of on-demand two-way interaction (initiated by the terminal, responded by the central terminal and written to the audit log) is still retained to generate virtual reference observations (VRS).

[0143] After passively receiving the correction packet, the inspection equipment first verifies the signature and checks its timeliness. Based on the approximate location of the equipment, it selects several grid points from nearby grid points and obtains the local correction at the equipment location through spatial interpolation (commonly bilinear or spatially correlated interpolation such as Kriging). Simultaneously, it synthesizes the correction covariance at the equipment location according to the error propagation rules (the interpolated covariance can be approximately expressed as the weighted sum of the covariances of each grid point plus the residual term of the interpolation model). The equipment projects this correction uncertainty into the observation space and merges it with the local observation noise model to construct an observation covariance / weight matrix for solution, so that the subsequent weighted solution can reflect the comprehensive confidence of the correction and the observation.

[0144] In observation preprocessing, the equipment performs cycle slip detection and rejection on local multi-frequency observations, and constructs ionospheric-free combinations as needed to reduce the influence of the first-order ionospheric term. Based on the observations corrected by interpolation, the equipment establishes linearized observation equations, generally written as...

[0145]

[0146] in The corrected observation residual vector (unit: meters). For continuous quantities to be estimated (such as position increments, tropospheric hanging parameters, etc.). This is the carrier phase integer ambiguity vector (unit: cycles). This is for observation noise (including the projected portion of the corrected uncertainty). and The design matrix is ​​derived from satellite geometry. Weighted least squares are used to find floating-point solutions to the above equations:

[0147]

[0148] in, weight matrix ( To observe the covariance (including local noise and corrected projection). Parameter description: —Observational covariance matrix (unit: The result is obtained by superimposing the receiver noise model and the interpolated corrected covariance. (Implicit in the algorithm) — Covariance matrix of floating-point ambiguity (unit: ).

[0149] After obtaining the floating-point ambiguity, the device uses a mature integerization method (such as LAMBDA) to first perform downcorrelation (integer Gaussian transform) on the floating-point ambiguity covariance to accelerate the search. It then searches for several optimal integer candidate solutions in the transform space and uses the cost ratio test and local consistency check to determine whether to accept the fixed solution. Mathematically, this means solving for the minimum objective... The device uses the ratio of the cost of the best candidate to the second-best candidate and the consistency of the residuals projected onto the location domain after the candidate is fixed as the passing threshold. If the ratio is greater than the preset threshold and the residual statistics (e.g., chi-square) are within the allowable range, the device will use integer ambiguity to fix and resolve, and use the fixed solution as the output result. Otherwise (insufficient ratio, abnormal residuals, expired correction packets, or downlink interruption), the device will not accept the fixation and will automatically trigger a degradation strategy.

[0150] A degradation strategy is automatically activated to ensure the continuity of the inspection task when reliable fixation fails. This strategy uses high-frequency pre-integration of the device's built-in inertial measurement unit (IMU) as short-time dynamic constraints, inter-frame feature displacements provided by visual odometry as relative closed-loop observations, and injects previously fixed and labeled reliable grid trajectories from historical periods as prior constraints into the fusion processor (which can be optimized using extended Kalman filtering or sliding window nonlinear least squares / factor graph). The output pose and its covariance are fused by weighting each input source according to its covariance. The rotation, velocity, and position increments generated by IMU pre-integration provide continuity for short-time motion; visual odometry provides drift suppression when texture / features are present; and historical grid priors can recover scale and absolute position references when both visual and inertial constraints are limited. The device attaches a clear accuracy confidence label to the fusion results (e.g., marked as "high / medium / low" by covariance principal value or custom level), and records the triggering reason, the serial number of the correction package used, the signature hash, and the quality indicators of all participating sensors to the local audit log. When communication is restored, the log is uploaded to the center for quality assessment and traceability.

[0151] In special or authorized scenarios, when one-way broadcasting is unavailable or the task requires higher precision, the device can send an authenticated request to the central terminal through a secure encrypted channel. After verifying the permissions, the central terminal can generate and sign the customized Virtual Reference Station Observation Data (VRS) according to the requested location. This interaction process is also fully recorded to meet audit and compliance requirements.

[0152] The entire process emphasizes three key engineering points: First, data traceability and security—all correction packages are timestamped, serial numbers are protected with asymmetric signatures, and the central end retains an audit chain; second, correction uncertainty is explicitly quantified and distributed along with the correction, and the terminal incorporates this uncertainty into the weight calculation during interpolation and solution to avoid erroneous confidence; third, it has an automatic and auditable degradation path—when ambiguity cannot be reliably fixed or correction fails, the terminal can locally fuse multiple sensors and output positioning results with confidence, ensuring the continuity and security of inspection operations.

[0153] Figure 2 This is a schematic diagram of a positioning system module based on BeiDou high-precision differential positioning provided in an embodiment of this application. Figure 2 The positioning system 10 based on BeiDou high-precision differential positioning shown includes at least the following parts: a central terminal 11 and a power grid inspection device 20.

[0154] In this embodiment, the central terminal 11 is used to receive observation data within the reference station network; obtain a state vector based on the observation data, the state vector including the reference station coordinate increment, tropospheric grid value, ionospheric grid value, and satellite clock bias; generate correction packets at regular grid points at preset intervals, the correction packets including the differential correction number and covariance description of the grid points; and send the correction packets to the equipment coverage area through a one-way broadcast channel to reduce the uplink exposure risk of equipment-side location information.

[0155] The central terminal 11 can be a positioning service platform deployed in an operation and maintenance private network or a data center / cloud platform. Specifically, it is implemented as a set of redundant hardware and software components, including a base station data access and management module (CORS access), a real-time intranet solver (supporting Kalman / weighted least squares and SSR fusion), a grid correction generation and management module, a digital signature and key management module (integrated with HSM or certificate management system), a one-way broadcast / downlink broadcast subsystem and an on-demand VRS generation subsystem, an audit and log management module, a terminal authentication and access control module, an operation and maintenance monitoring and alarm module, and external interfaces (API / NTRIP / industry protocol adapter). In terms of physical configuration, the central terminal 11 can be deployed as a local data center cluster, edge computing node, or a hybrid cloud deployment, employing high availability redundancy and backup strategies to ensure service continuity. For security, the central terminal 11 is responsible for hashing and asymmetric signing of correction packets (private keys are stored in the HSM), maintaining terminal certificates and access control policies, and managing control plane transactions (certificate issuance, maintenance commands, on-demand VRS requests, etc.) through a controlled bidirectional channel. The data plane can use one-way downlink broadcast or read-only multicast channels to achieve "one-way service." The central terminal 11 also provides an operation and maintenance console and interface, allowing maintenance personnel to view the base station status, grid coverage quality, signature / audit records, and terminal health status. It can also dynamically adjust grid density and correction distribution strategies based on ionospheric activity and terrain features. Please refer to the following for details. Figure 1 The details and their corresponding descriptions are not repeated here.

[0156] In this embodiment, the power grid inspection device 20 receives correction packets and verifies their signatures and timeliness; performs interpolation on four neighboring grid points to obtain the covariance after correction and propagation at the device location; and constructs an observation weight matrix. It performs cycle slip detection and ionosphere-free combination processing on local observations, constructs a linearized observation equation to obtain a floating-point solution, and applies the LAMBDA method to search for integer ambiguities. A preset ratio threshold is used as the criterion for successful ambiguity fixation. The residual is subjected to a chi-square test for integrity monitoring. If the ambiguity is successfully fixed and passes the integrity monitoring, the location result is obtained by fixing and resolving based on the integer ambiguity. Please refer to the following for details. Figure 1 The details and their corresponding descriptions are not repeated here.

[0157] like Figure 2As shown, the positioning system 10 based on BeiDou high-precision differential positioning can include multiple power grid inspection devices 20. Specifically, the power grid inspection device 20 can be a UAV-borne high-precision GNSS receiver, an overhead line inspection vehicle or its supporting ground inspection vehicle terminal, a tower-climbing robot, a handheld inspection terminal, a fixed online monitoring node, or an embedded edge terminal, etc. Each power grid inspection device 20 typically integrates a multi-system multi-frequency GNSS receiving module (supporting BeiDou / GPS / multi-frequency calculation), an inertial measurement unit (IMU), a visual sensor (camera / stereo camera / depth camera), an optional LiDAR, a communication unit (e.g., private 4G / 5G network, dedicated short-range wireless or broadcast receiver), a local storage and security chip (e.g., TPM or secure element), and a software module for local calculation and recording audit logs. The power grid inspection equipment 20 is responsible for passively receiving and verifying grid correction packets from the central end, performing local interpolation and covariance propagation, performing floating-point solutions and integerization attempts, triggering local degradation fusion as needed, and generating confident positioning output and integrity / audit records.

[0158] Figure 3 This is a power grid inspection device 20 provided in one embodiment of this application. For example... Figure 3 As shown, the power grid inspection equipment 20 includes at least the following components: processor 21 and memory 22.

[0159] In this embodiment, the memory 22 is used to store executable instructions of the processor 21, which, when configured to execute instructions, implement... Figure 1 The positioning method based on BeiDou high-precision differential positioning is shown in the figure.

[0160] It is understood that the positioning method, system, and equipment based on BeiDou high-precision differential positioning provided in this application uniformly calculates and generates gridded differential correction information from the reference station network observation data at the central end, and then transmits the correction data to the power grid inspection equipment through a one-way broadcast method. This allows the terminal to obtain high-precision positioning services without continuously uploading location information, thereby improving positioning accuracy and real-time performance while reducing the risk of equipment-side location information exposure and enhancing the system's data security and traceability. Simultaneously, by introducing covariance description into the correction data and combining it with terminal-side interpolation calculation and weight construction mechanisms, observation quality can be dynamically modeled, thereby improving the stability and reliability of positioning calculations. Furthermore, by combining ambiguity fixing judgment with integrity monitoring mechanisms, the reliability of positioning results can be effectively guaranteed. In the event of signal obstruction or communication anomalies, adaptive degradation of the positioning mode can be achieved through multi-source information fusion, ensuring the continuous positioning capability of the inspection equipment in complex environments, thereby comprehensively improving the safety, stability, and intelligence level of power grid inspection operations.

[0161] Those skilled in the art should recognize that the above embodiments are only used to illustrate this application and are not intended to limit this application. Any appropriate changes and variations made to the above embodiments within the essential spirit and scope of this application fall within the scope of protection claimed in this application.

Claims

1. A positioning method based on BeiDou high-precision differential positioning, applied to a BeiDou positioning network consisting of power grid inspection equipment and a central terminal, characterized in that, The method includes: The central terminal receives observation data from the reference station network; A state vector is obtained based on the observation data, and the state vector includes the base station coordinate increment, tropospheric grid value, and ionospheric grid value. Correction packages are generated at preset intervals for regular grid points, and the correction packages include differential corrections and covariance descriptions for the grid points; The correction package is sent to the device coverage area through a one-way broadcast channel to reduce the risk of uplink exposure of device-side location information; The power grid inspection equipment receives the correction packet and performs interpolation operations on the four neighboring grid points according to the correction packet to obtain the correction number at the equipment and the covariance after propagation; Construct an observation weight matrix based on the corrections at the device and the propagated covariance; The power grid inspection equipment performs cycle slip detection and ionosphere-free combination processing on local observations, constructs linearized observation equations to obtain floating-point solutions, and applies the LAMBDA method to search for integer ambiguities. A preset ratio threshold is used as the criterion for successful ambiguity fixation. Perform a chi-square test on the solution residuals to monitor integrity. If the ambiguity is successfully fixed and passes the integrity monitoring, then fix and resolve based on the integer ambiguity to obtain the positioning result.

2. The positioning method based on BeiDou high-precision differential positioning according to claim 1, characterized in that, The step of obtaining the state vector based on the observed data includes: Adaptive filtering algorithms or weighted least squares methods are applied to perform quality control and systematic error removal on the raw observations of the reference station network in order to estimate the state vector; When generating the correction package, the calculated correction covariance description and grid coordinate information are encapsulated, and the data digest is signed using an asymmetric encryption algorithm.

3. The positioning method based on BeiDou high-precision differential positioning according to claim 2, characterized in that, Interpolation is performed on the four neighboring grid points to obtain the device-corrected and propagated covariance, and the observation weight matrix is ​​constructed, including: Based on the approximate current location of the power grid inspection equipment, a preset number of grid points are selected in the vicinity, and bilinear interpolation or spatial correlation interpolation algorithm is executed to extract the delay correction term of the location of the power grid inspection equipment. The covariance description of the grid points is transmitted to the power grid inspection equipment by applying the error propagation law, and the weight matrix that can dynamically reflect the different satellite observation quality is constructed by combining the local preset random model and the residual error of the interpolation model.

4. The positioning method based on BeiDou high-precision differential positioning according to claim 3, characterized in that, The method of applying LAMBDA to search for integer ambiguity, using a preset ratio threshold as the criterion for successful ambiguity fixation, includes: Construct a combined observation system of wide and narrow alleyways; The original ambiguity covariance matrix is ​​decorrelated by applying integer Gaussian transform, which maps the search space to a transform space, and the optimal integer candidate solution is determined in the transform space. The conditions for determining ambiguity fixation failure include: the ratio of the optimal candidate solution to the second-best candidate solution does not reach the preset safety threshold, or the chi-square test statistic of the calculated residual components exceeds the limit corresponding to the significance level; when the above failure conditions, communication link interruption, or grid data timeliness failure is detected, the positioning mode is triggered to switch autonomously from carrier phase differential mode to degraded fusion mode.

5. The positioning method based on BeiDou high-precision differential positioning according to claim 1, characterized in that, The method further includes: If the determination of ambiguity fixation failure, integrity monitoring abnormality, or communication interruption occurs, a downgrade fusion strategy will be triggered. The degradation fusion strategy includes outputting positioning results with accuracy confidence labels by fusing inertial measurement unit pre-integration, visual odometry, and historical grid trajectories.

6. The positioning method based on BeiDou high-precision differential positioning according to claim 5, characterized in that, The method further includes: The inertial measurement unit built into the power grid inspection equipment is invoked to perform high-frequency motion pre-integration, and relative pose estimation is performed by combining the displacement of feature points captured by the visual sensor. The fixed grid trajectory within the historical period is injected into the fusion filter as a priori motion constraint to compensate for positioning drift in signal occlusion environments. The output includes a continuous stream of location information, including the current solution status, estimation accuracy, availability classification labels, and anomaly audit logs.

7. The positioning method based on BeiDou high-precision differential positioning according to claim 1, characterized in that, The method further includes: The central end dynamically adjusts the grid coverage density of the local area based on the regional ionospheric activity index and topographic relief characteristics, and uses a polynomial fitting model to extract the large-scale atmospheric delay gradient coefficient and encapsulate it in the correction package. When performing interpolation, the power grid inspection equipment performs nonlinear error compensation on the linear interpolation results based on the coefficients of the polynomial fitting model, so as to reduce the non-modeling impact of atmospheric delay residue on carrier phase observations.

8. The positioning method based on BeiDou high-precision differential positioning according to claim 7, characterized in that, The method further includes: When the power grid inspection equipment identifies that the quality of the one-way broadcast signal is lower than the preset limit or is in a specific authorized task area, it requests information from the uplink location of the central terminal through a secure encrypted channel. In response to the request, the central terminal calculates and generates virtual reference station observation data for the corresponding location, and feeds back customized differential correction parameters through a two-way communication channel; The central terminal is also used to write the complete interaction process, including identity tokens, request parameters, and distribution records, into the audit log.

9. A positioning system based on BeiDou high-precision differential positioning, characterized in that, The system includes: The central terminal is used to receive observation data from the reference station network; obtain a state vector based on the observation data, the state vector including the reference station coordinate increment, tropospheric grid value, and ionospheric grid value; generate correction packets at regular grid points at preset intervals, the correction packets including the differential correction number and covariance description of the grid points; and send the correction packets to the equipment coverage area through a one-way broadcast channel to reduce the uplink exposure risk of equipment-side location information. The power grid inspection equipment is used to receive the correction packet, perform interpolation on the four neighboring grid points to obtain the correction number at the equipment location and the propagated covariance, construct an observation weight matrix based on the correction number at the equipment location and the propagated covariance, perform cycle slip detection and ionospheric-free combination processing on the local observations, construct a linearized observation equation to obtain a floating-point solution, and apply the LAMBDA method to search for integer ambiguities, using a preset ratio threshold as the criterion for successful ambiguity fixation; perform a chi-square test on the solution residuals for integrity monitoring, and if the ambiguity is successfully fixed and passes the integrity monitoring, then fix and resolve based on the integer ambiguity to obtain the positioning result.

10. A power grid inspection device, characterized in that, include: processor; as well as A memory having computer-readable instructions stored thereon for controlling the processor to execute the positioning method based on BeiDou high-precision differential positioning as described in any one of claims 1 to 8.

Images