An underground cable positioning method based on composite electromagnetic field and multi-source information fusion

Abstract

This invention discloses an underground cable location method based on composite electromagnetic field and multi-source information fusion, comprising the following steps: S1, using an AC-DC coupled current generator to synchronously inject a large ampere DC current and a high-frequency small current into the target buried cable to form a magnetic field in the space surrounding the target buried cable; S2, acquiring the DC stable magnetic field signal and the high-frequency alternating magnetic field signal to obtain the amplitude of the DC stable magnetic field signal and the triaxial characteristics of the high-frequency alternating magnetic field; S3, performing Kalman filtering analysis based on the amplitude of the DC stable magnetic field signal to obtain an estimate of the lateral offset of the target buried cable; S4, obtaining an estimate of the direction of the target buried cable based on the triaxial characteristics of the high-frequency alternating magnetic field; S5, obtaining a location dataset based on the lateral offset estimate and the direction estimate, and obtaining the cable center path based on the location dataset; S6, extracting the cable burial depth data of the target buried cable and outputting the location information of the target buried cable.

Description

Technical Field

[0001] This invention relates to the field of underground cable location technology, and in particular to an underground cable location method based on the fusion of composite electromagnetic fields and multi-source information. Background Technology

[0002] With the large-scale construction and grid connection of new energy power plants, the laying density and network complexity of their underground cables have increased significantly. The concealed nature of cable routes, multi-layered cross-layouts, and uneven soil conditions make accurate cable detection, path tracking, and burial depth determination key challenges in operation and maintenance, fault diagnosis, and construction avoidance. Traditional positioning methods relying on manual experience, cable identification instruments, or single physical field measurements often exhibit poor positioning continuity, insufficient accuracy, and weak anti-interference capabilities when faced with strong electromagnetic interference, complex geological conditions, and environments with dense metal components. These methods fail to meet the practical needs of new energy power plants for digitized cable assets and refined operation and maintenance.

[0003] Currently, detection technologies for buried cables mainly rely on single physical field principles. For example, detection methods based on low-frequency or power-frequency electromagnetic induction locate cables by detecting the magnetic field generated by the cable's own current or induced current. However, the signal is weak in passive or weak current conditions and is easily affected by coupling interference from surrounding parallel cables or metal pipelines. Detection methods based on DC or pulsed magnetic fields inject a specific current into the cable and measure the resulting static or transient magnetic field. While these methods can effectively identify target cables, their signals attenuate rapidly in conductive soil, and the measurement data is easily distorted in the presence of ferromagnetic materials or areas with strong geomagnetic anomalies. In addition, other modal methods such as ground-penetrating radar and distributed fiber optic sensing, while providing supplementary information, have limitations such as sensitivity to scattering from metallic targets, complex signal interpretation, susceptibility to environmental temperature and vibration noise, high equipment cost, and inconvenient deployment.

[0004] While existing technologies have attempted to employ multi-sensor combinations for detection, they typically suffer from the following shortcomings: First, multiple signals often employ time-division injection or sequential acquisition modes, leading to temporal asynchrony and spatial registration errors between data from different physical fields, making true synchronous fusion difficult. Second, there is a lack of effective anti-interference design between sensors; for example, DC magnetic field measurements are susceptible to AC magnetic field interference, while high-frequency magnetic field measurements may be affected by geomagnetic background noise, resulting in signal crosstalk and a decrease in signal-to-noise ratio. Third, data processing often relies on a single algorithm or simple weighting, failing to fully integrate the mechanistic characteristics of different physical fields (such as the spatial gradient stability of DC magnetic fields and the phase directionality of high-frequency magnetic fields) for deep coupling and collaborative analysis, which can easily lead to problems such as positioning jumps and discontinuous path fitting in complex environments.

[0005] Therefore, there is an urgent need for an underground cable positioning method that can simultaneously acquire multiple complementary magnetic field information, has strong anti-interference capabilities, and achieves high-precision fusion of multi-source data. Summary of the Invention

[0006] In view of the above-mentioned prior art, the present invention provides an underground cable positioning method based on the fusion of composite electromagnetic field and multi-source information, which mainly solves the technical problems existing in the background art.

[0007] To achieve the above objectives, the technical solution of the present invention is implemented as follows: A method for locating underground cables based on the fusion of composite electromagnetic fields and multi-source information includes the following steps: Step S1: Simultaneously inject a large ampere DC current and a high-frequency small current into the target buried cable using an AC-DC coupling current generator to form a magnetic field in the space around the target buried cable. Step S2: Use a multi-sensor collaborative acquisition device to acquire the DC stable magnetic field signal and the high-frequency alternating magnetic field signal of the magnetic field, and obtain the amplitude of the DC stable magnetic field signal and the triaxial characteristics of the high-frequency alternating magnetic field based on the DC stable magnetic field signal and the high-frequency alternating magnetic field signal, respectively. Step S3: Perform Kalman filtering analysis based on the amplitude of the DC stable magnetic field signal to obtain an estimate of the lateral offset of the target buried cable; Step S4: Estimate the direction of the target buried cable based on the triaxial characteristics of the high-frequency alternating magnetic field; Step S5: Based on the lateral offset estimate and the direction estimate, obtain the positioning dataset, and obtain the cable center path based on the positioning dataset; Step S6: Extract the burial depth data of the target underground cable, and output the positioning information of the target underground cable by combining the cable center path and the burial depth data.

[0008] Optionally, the AC-DC coupling current generating device includes: a high-power DC power supply, a frequency converter signal source, a coupler, a power amplifier, and an AC-DC coupling output module; The output terminal of the high-power DC power supply is electrically connected to the input terminal of the coupler, the frequency converter signal source is electrically connected to the input terminal of the coupler through the power amplifier, and the output terminal of the coupler is electrically connected to the input terminal of the AC-DC coupling output module.

[0009] Optionally, the multi-sensor collaborative acquisition device includes: two sets of magnetometers, two sets of triaxial high-frequency coils, a follower, a frequency selector, an A / D converter, an amplifier, a CAN bus, and a CPU; The two sets of magnetometers are used to collect the DC stable magnetic field signal generated by the target buried cable. The signal output terminals of the two sets of magnetometers are connected to the signal input terminal of the CPU through the CAN bus. The amplitude of the DC stable magnetic field signal is obtained after processing by the CPU. The two sets of triaxial high-frequency coils are used to acquire the high-frequency alternating magnetic field signal generated by the target buried cable. The signal output terminals of the two sets of triaxial high-frequency coils are connected in series with the follower, the frequency selector, the amplifier, and the input terminal of the A / D converter. The output terminal of the A / D converter is connected to another signal input terminal of the CPU. The high-frequency alternating magnetic field triaxial characteristics are obtained after processing by the CPU.

[0010] Optionally, step S3 includes: Step S31: Set up stations along the survey line. Construct a state vector using the lateral offset of the target buried cable relative to the center of the survey line at each station and the rate of change of offset along the direction of the survey line station. The expression is:

[0011] in, Let s be the state vector of station s. Let be the rate of change of offset at station s. Estimate the lateral offset at station s; Step S32: Based on the amplitude of the DC stable magnetic field signal from the two sets of magnetometers, obtain the DC differential measured observation, expressed as:

[0012] in, For DC differential measured observations at station s, The amplitude of the stable DC magnetic field signal obtained by the electromagnetic meter located to the left of the target buried cable. The amplitude of the DC stable magnetic field signal obtained by the electromagnetic meter located on the right side of the target buried cable; Step S33: Construct a monotonic mapping function between the lateral offset and the DC magnetic field difference based on the Biot-Savart law, and construct the observation equation according to the monotonic mapping function, the expression of which is:

[0013]

[0014] in, Let s be the state vector of station s. Here is the state transition matrix. Let f be the state vector of the previous station. This is the monotonic mapping function between the DC magnetic field difference and the lateral offset, obtained according to the Biot-Savart law. To inject direct current, As the initial value for burial depth, The distance between the two sets of magnetometers is half-pitch. For noise in the offset change rate process, To observe Gaussian noise, The predicted observations at station s are obtained based on the lateral offset solution. Step S34: Traverse all measurement stations along the survey line in sequence, and perform Kalman filter prediction-update recursive estimation for each station in sequence to obtain the lateral offset estimate of the target buried cable.

[0015] Optionally, step S34 includes: The predicted state vector of the current station is obtained by predicting the current station using the posterior state vector and state transition matrix of the previous station. The expression is as follows:

[0016] in, Predict the state vector for the current site. Let be the posterior state vector of the previous station. This is the state transition matrix; The prior prediction covariance of the current station is calculated using the posterior covariance of the previous station, expressed as:

[0017] in, For the prior prediction covariance of the current site, For the posterior covariance of the previous site, For process noise covariance; The Kalman gain is calculated based on the prior prediction covariance of the current site, and the expression is as follows:

[0018] in, For Kalman gain, For Jacobian matrices, To fix the observation noise covariance; Calculate the observation residual between the measured and predicted DC differential observations, and combine it with the Kalman gain to correct the current station's predicted state, thus obtaining the optimal posterior predicted state for the current station. The expression is as follows:

[0019] in, This represents the optimal posterior state vector for the current site. For the predicted observations obtained from the prior predicted state, To observe the residuals, These are the DC differential measured observations at station s; The posterior covariance of the current site is updated synchronously using the following expression:

[0020] in, The posterior covariance of the current site. It is an identity matrix.

[0021] Optionally, step S4 includes: The triaxial characteristics of the high-frequency alternating magnetic field include: a triaxial amplitude vector and a phase vector; Step S41: Calculate the composite amplitude and its direction angle in the horizontal plane based on the three-axis amplitude vector. Based on the direction angle, obtain two candidate orientation angles for the current station, expressed as:

[0022]

[0023] Among them, A h For the synthesized amplitude, A x A represents the X-axis magnitude in the three-axis magnitude vector. y θ represents the Y-axis magnitude in the three-axis magnitude vector. h θ is the direction angle. c To move towards candidate angles; Step S42: Based on the two candidate orientation angles of the current station and the effective orientation of the previous station, determine the effective orientation candidate value of the current station, and traverse each station in turn to obtain the effective orientation candidate value of each station of the target buried cable. Step S43: Process the valid candidate values ​​of each station to obtain the trajectory of the target buried cable, wherein the trajectory includes the trajectory estimate of each station.

[0024] Optionally, step S5 includes: Step S51: Based on the lateral offset estimate and the orientation estimate, construct a positioning dataset, which includes: the station coordinates, lateral offset estimate, orientation estimate, DC component evaluation index, and high frequency component evaluation index for each station. Step S52: Using the spatial coordinates of each measurement station as a unified index, perform spatial serialization processing on the positioning dataset corresponding to each station to generate an ordered station sequence with coordinates increasing along the survey line. Step S53: Determine the center point of the survey line. Using the center point of the survey line as the spatial projection reference, perform spatial projection calculations by combining the lateral offset estimation and the direction estimation to obtain a set of candidate cable center points. Step S54: Read the DC component evaluation index and high frequency component evaluation index of each station sequentially along the ordered station sequence, and select the first station in the ordered station sequence whose DC component evaluation index and high frequency component evaluation index both meet the preset confidence threshold as the initial seed point. Step S55: Based on the initial seed point and the set of candidate cable center points, execute a combined strategy of ridge tracing and local path fitting to obtain the cable center path.

[0025] Optionally, step S55 includes: Using the estimated direction from the initial seed point as the first-round tracking direction, adjacent stations are located; Select the cable center candidate points corresponding to adjacent stations from the set of cable center candidate points; Construct a spatial sliding window centered on the candidate point at the center of the cable and containing several neighboring candidate points; Within the sliding window, the least squares method is used to fit a curve to the set of all candidate points within the sliding window. The coordinates and tangent direction of the fitted curve at the current position are solved. Based on the solution results, the original candidate points for the cable center are updated and corrected to obtain the corrected candidate points. The modified candidate points are determined as cable center ridge nodes and stored in the complete cable center path; The corrected candidate point is set as the initial seed point for the next round of tracing iteration, and the iteration continues to obtain the complete cable center path.

[0026] Optionally, step S6 involves extracting the burial depth data of the target underground cable, including: Step S61: Construct a model of the relationship between magnetic field characteristics and burial depth based on the Biot-Savart law, the expression of which is:

[0027] in, For vertical burial depth estimation, The characteristic amplitude of the magnetic field. To inject current, For lateral offset estimation, Permeability of free space; Step S62: Obtain the amplitude of the DC stable magnetic field signal and the characteristic amplitude of the high-frequency magnetic field at the current site. Based on the amplitude of the DC stable magnetic field signal, the characteristic amplitude of the high-frequency magnetic field, and the relationship model, calculate the estimated value of the DC source burial depth and the estimated value of the high-frequency source burial depth. Step S63: Calculate the initial burial depth fusion value of the current site by weighted averaging the DC source burial depth estimate and the high-frequency source burial depth estimate. Step S64: Repeat steps S61-S63 to obtain the initial burial depth fusion values ​​for all stations; Step S65: Identify anomalies in the initial burial depth fusion values ​​of all sites, correct the identified initial burial depth fusion values, and obtain continuous cable burial depth data for the target underground cable.

[0028] The beneficial effects of this invention are as follows: This invention provides an underground cable positioning method based on the fusion of composite electromagnetic fields and multi-source information. By acquiring the amplitude of the DC stable magnetic field signal and the triaxial characteristics of the high-frequency alternating magnetic field around the target buried cable, the lateral offset and direction of the target buried cable are estimated, thereby obtaining a continuous and complete cable center path. Furthermore, the burial depth data of the target buried cable is extracted, and finally, the positioning information of the target buried cable is output. This method can suppress single-point observation errors and overcome the problems of weak anti-interference ability and poor positioning accuracy of traditional single magnetic field detection methods in complex pipeline environments, achieving high-precision detection of the target buried cable path, lateral offset, and burial depth. Attached Figure Description

[0029] Figure 1 This is a flowchart illustrating an underground cable location method based on the fusion of composite electromagnetic fields and multi-source information provided in an embodiment of the present invention. Figure 2 This is a structural diagram of the AC-DC coupling current generating device provided in an embodiment of the present invention; Figure 3 This is a structural diagram of the multi-sensor collaborative acquisition device provided in an embodiment of the present invention; Figure 4 This is a flowchart illustrating step S3 provided in an embodiment of the present invention; Figure 5 This is a flowchart illustrating step S5 provided in an embodiment of the present invention; Figure 6 This is a schematic diagram of the process for extracting the cable burial depth data of the target underground cable in step S6 of the embodiment of the present invention. Detailed Implementation

[0030] To fully understand this invention, a detailed structure will be presented in the following description to illustrate the technical solution proposed by this invention. Optional embodiments of the invention are described in detail below; however, in addition to these detailed descriptions, the invention may have other embodiments.

[0031] Example Please refer to the attached document. Figure 1 This application provides a method for locating underground cables based on the fusion of composite electromagnetic fields and multi-source information, including the following steps: Step S1: Simultaneously inject a large ampere DC current and a high-frequency small current into the target buried cable using an AC-DC coupling current generator to form a magnetic field in the space around the target buried cable. Step S2: Use a multi-sensor collaborative acquisition device to acquire the DC stable magnetic field signal and the high-frequency alternating magnetic field signal of the magnetic field, and obtain the amplitude of the DC stable magnetic field signal and the triaxial characteristics of the high-frequency alternating magnetic field based on the DC stable magnetic field signal and the high-frequency alternating magnetic field signal, respectively. Step S3: Perform Kalman filtering analysis based on the amplitude of the DC stable magnetic field signal to obtain an estimate of the lateral offset of the target buried cable; Step S4: Estimate the direction of the target buried cable based on the triaxial characteristics of the high-frequency alternating magnetic field; Step S5: Based on the lateral offset estimate and the direction estimate, obtain the positioning dataset, and obtain the cable center path based on the positioning dataset; Step S6: Extract the burial depth data of the target underground cable, and output the positioning information of the target underground cable by combining the cable center path and the burial depth data.

[0032] Specifically, an AC-DC coupled current generator is used to create a magnetic field around the target buried cable. A multi-sensor collaborative acquisition device is then used to obtain the amplitude of the stable DC magnetic field signal and the triaxial characteristics of the high-frequency alternating magnetic field. To better locate the target buried cable, Kalman filtering analysis is performed based on the amplitude of the stable DC magnetic field signal to estimate the lateral offset of the target buried cable. The direction of the target buried cable is estimated based on the triaxial characteristics of the high-frequency alternating magnetic field. Based on the lateral offset and direction estimates, the discrete site data is processed to obtain a continuous and complete cable center path. To more accurately locate the target buried cable, the cable burial depth data is extracted from the location dataset. The location information of the target buried cable is obtained based on the lateral offset estimate, direction estimate, and cable burial depth data.

[0033] Since the DC stable magnetic field signal and the high-frequency alternating magnetic field signal are physically independent, they do not interfere with each other when acquiring the two types of magnetic field signals. This enables the spatiotemporal consistent acquisition and deep fusion of multi-source magnetic field data, significantly improving the continuity, accuracy, and reliability of underground cable positioning in complex geological and strong interference environments. Next, the lateral offset and direction estimation of each station are solved, and a combination strategy of ridge tracking and local path fitting is adopted to achieve the joint information of multiple stations. The cable burial depth data of the target underground cable is extracted, and the positioning information of the target underground cable is output. This method can suppress single-point observation errors and overcome the problems of weak anti-interference ability and poor positioning accuracy of traditional single magnetic field detection methods in complex pipeline environments, realizing high-precision detection of the target underground cable path, lateral offset, and burial depth.

[0034] As an optional implementation method, please refer to the appendix. Figure 2 The AC-DC coupling current generating device includes: a high-power DC power supply, a frequency converter signal source, a coupler, a power amplifier, and an AC-DC coupling output module; The output terminal of the high-power DC power supply is electrically connected to the input terminal of the coupler, the frequency converter signal source is electrically connected to the input terminal of the coupler through the power amplifier, and the output terminal of the coupler is electrically connected to the input terminal of the AC-DC coupling output module.

[0035] Specifically, the AC-DC coupling current generator has the function of precise adjustment of signal amplitude and frequency; among them, the high-power DC power supply is the core component of the DC signal source, which outputs a large ampere DC current with an amplitude of 10-50A. When this current flows through the target buried cable conductor, it generates a strong and stable DC magnetic field in the space around the cable, which is used to accurately capture the spatial orientation deviation of the cable. The frequency converter outputs a small-amplitude, high-frequency, low-current signal with a frequency of 10-100kHz and an amplitude of 1-5mA. After the amplitude is amplified by the power amplifier, the signal is transmitted to the coupler and coupled with the DC current output by the high-power DC power supply. The coupled mixed signal is processed by the AC-DC coupling output module and then stably injected into one or both ends of the target buried cable. Finally, a low-intensity, periodically changing high-frequency alternating magnetic field is formed around the cable, which is used to quickly lock the direction of the cable.

[0036] By coordinating the components of the AC-DC coupling current generator, a composite electromagnetic field environment of "strong stable magnetic field and weak alternating magnetic field" is constructed through the synchronous injection of large ampere DC current and high-frequency small current. The two magnetic fields act independently on the space around the cable based on different current excitation characteristics and do not interfere with each other. This method can make full use of the differences in physical characteristics of the two magnetic fields to provide complementary observation data sources for subsequent multi-sensor acquisition. At the same time, synchronous triggering ensures the spatiotemporal consistency of signal injection and acquisition, providing a basic spatiotemporal reference for subsequent cross-modal data fusion.

[0037] It should be noted that the coupler has an adaptive impedance matching function, which can dynamically adjust its matching parameters according to the characteristic impedance of the target buried cable to ensure that there is no amplitude attenuation or phase distortion during the coupling process of DC current signal and high-frequency current signal, thus ensuring signal injection efficiency.

[0038] As an optional implementation method, please refer to the appendix. Figure 3 The multi-sensor collaborative acquisition device includes: two sets of magnetometers, two sets of triaxial high-frequency coils, a follower, a frequency selector, an A / D converter, an amplifier, a CAN bus, and a CPU; The two sets of magnetometers are used to collect the DC stable magnetic field signal generated by the target buried cable. The signal output terminals of the two sets of magnetometers are connected to the signal input terminal of the CPU through the CAN bus. The amplitude of the DC stable magnetic field signal is obtained after processing by the CPU. The two sets of triaxial high-frequency coils are used to acquire the high-frequency alternating magnetic field signal generated by the target buried cable. The signal output terminals of the two sets of triaxial high-frequency coils are connected in series with the follower, the frequency selector, the amplifier, and the input terminal of the A / D converter. The output terminal of the A / D converter is connected to another signal input terminal of the CPU. The high-frequency alternating magnetic field triaxial characteristics are obtained after processing by the CPU.

[0039] Specifically, two sets of magnetometers are used to collect the DC stable magnetic field signal around the target buried cable. The collection frequency is not less than 1kHz to ensure the integrity of the DC stable magnetic field signal collection. The two sets of magnetometers are equipped with built-in filtering modules, which can remove the interference of high-frequency alternating magnetic field signals through filtering, accurately capture the DC stable magnetic field signal, and then transmit the collected DC stable magnetic field signal to the CPU through the CAN bus to obtain the DC stable magnetic field signal amplitude. The triaxial high-frequency coil is used to collect the high-frequency alternating magnetic field signal (triaxial component signal) around the target buried cable. The acquisition frequency is not less than 500kHz. The acquired high-frequency alternating magnetic field signal is sequentially impedance matched and buffered by a follower. The frequency selector filters out the high-frequency signal in the target frequency range (10-100kHz). The amplifier amplifies the signal. The A / D converter converts the amplified high-frequency signal into a digital signal. The CPU processes the digital signal to obtain the triaxial characteristics of the high-frequency alternating magnetic field. The CPU receives the DC stable magnetic field signal collected by the magnetometer and the high-frequency alternating magnetic field signal collected by the triaxial high-frequency coil. While collecting data, the CPU records the spatial position coordinates (longitude, latitude, and elevation) of the magnetometer and the triaxial high-frequency coil in real time through the built-in GPS / BeiDou positioning module, and simultaneously marks the timestamp of each set of collected data. According to the on-site detection accuracy requirements, the spacing between the measuring lines is set to no more than 0.4m, and the station step is set to no more than 0.1m to ensure the spatial continuity of the collected data.

[0040] It should be noted that the multi-sensor collaborative acquisition device is deployed in a standardized manner along the surface survey line of the target buried cable detection corridor. Specifically, two sets of magnetometers are symmetrically positioned on either side of the center of the survey line, with a spacing of 0.1-0.3m. This spacing ensures that when there is a deviation between the magnetometers and the cable's center, the two sets of devices can acquire significantly different stable DC magnetic field signals. Two sets of triaxial high-frequency coils are positioned in front of and behind the magnetometer sets, with a spacing of 0.2-0.5m. These are used to acquire high-frequency alternating magnetic field signals. When the coils deviate from the cable center, the high-frequency signal readings will show a regular deviation, which can help determine the cable's direction. The two sets of magnetometers and the two sets of high-frequency coils can be integrated and fixed on handheld detection devices or mounted on mobile carts, enabling portable and flexible detection or automated line-scanning detection, adapting to the detection needs of different scenarios such as urban core areas and new energy power stations.

[0041] Since the DC stable magnetic field does not change with time, it cannot induce an electromotive force in the coil. Therefore, the signal acquired by the triaxial high-frequency coil is naturally unaffected by the DC stable magnetic field, enabling accurate separation and acquisition of the two magnetic field signals.

[0042] As an optional implementation method, please refer to the appendix. Figure 4 Step S3 includes: Step S31: Set up stations along the survey line. Construct a state vector using the lateral offset of the target buried cable relative to the center of the survey line at each station and the rate of change of offset along the direction of the survey line station. The expression is:

[0043] in, Let s be the state vector of station s. Let be the rate of change of offset at station s. Estimate the lateral offset at station s; Step S32: Based on the amplitude of the DC stable magnetic field signal from the two sets of magnetometers, obtain the DC differential measured observation, expressed as:

[0044] in, For DC differential measured observations at station s, The amplitude of the stable DC magnetic field signal obtained by the electromagnetic meter located to the left of the target buried cable. The amplitude of the DC stable magnetic field signal obtained by the electromagnetic meter located on the right side of the target buried cable; Step S33: Construct a monotonic mapping function between the lateral offset and the DC magnetic field difference based on the Biot-Savart law, and construct the observation equation according to the monotonic mapping function, the expression of which is:

[0045]

[0046] in, Let s be the state vector of station s. Here is the state transition matrix. Let f be the state vector of the previous station. This is the monotonic mapping function between the DC magnetic field difference and the lateral offset, obtained according to the Biot-Savart law. To inject direct current, As the initial value for burial depth, The distance between the two sets of magnetometers is half-pitch. For noise in the offset change rate process, To observe Gaussian noise, The predicted observations at station s are obtained based on the lateral offset solution. Step S34: Traverse all measurement stations along the survey line in sequence, and perform Kalman filter prediction-update recursive estimation for each station in sequence to obtain the lateral offset estimate of the target buried cable.

[0047] Specifically, a state vector is constructed, which describes the lateral position and positional change trend of the target buried cable at the station, providing a basis for the Kalman filter prediction stage. Two sets of magnetometers are symmetrically set on both sides of the center of the survey line. The DC differential value is obtained by subtracting the amplitude of the DC stable magnetic field signal from the two sets of magnetometers. This DC differential value is used as the DC differential measured observation, which provides a basis for subsequent Kalman filter updates. Then, the state transition equation and observation equation are constructed. Based on the Biot-Savart law, a mapping function is obtained. The observation equation uses the mapping function to calculate the predicted observation value, which facilitates the calculation of the residual in the subsequent Kalman filter update stage. Then, Kalman filter prediction-update recursive estimation is performed station by station, outputting the optimal posterior state vector (lateral offset estimation), observation residual, and posterior covariance for each station. The magnitude of the posterior covariance characterizes the reliability of the estimation result. This posterior covariance can be directly used to evaluate the reliability of the offset in subsequent evaluations.

[0048] As an optional implementation, step S34 includes: The predicted state vector of the current station is obtained by predicting the current station using the posterior state vector and state transition matrix of the previous station. The expression is as follows:

[0049] in, Predict the state vector for the current site. Let be the posterior state vector of the previous station. This is the state transition matrix; The prior prediction covariance of the current station is calculated using the posterior covariance of the previous station, expressed as:

[0050] in, For the prior prediction covariance of the current site, For the posterior covariance of the previous site, For process noise covariance; The Kalman gain is calculated based on the prior prediction covariance of the current site, and the expression is as follows:

[0051] in, For Kalman gain, For Jacobian matrices, To fix the observation noise covariance; Calculate the observation residual between the measured and predicted DC differential observations, and combine it with the Kalman gain to correct the current station's predicted state vector, thus obtaining the optimal posterior predicted state vector for the current station. The expression is as follows:

[0052] in, This represents the optimal posterior state vector for the current site. The predicted observations are obtained from the prior predicted state vector. To observe the residuals, These are the DC differential measured observations at station s; The posterior covariance of the current site is updated synchronously using the following expression:

[0053] in, The posterior covariance of the current site. It is an identity matrix.

[0054] Specifically, the lateral offset of the target buried cable is estimated by sequentially performing Kalman filtering prediction-update recursive estimation on each station. First, based on the continuity of the cable route, the prior value of the current station's offset is predicted from the state of the previous station. Then, the optimal posterior state vector is obtained by combining DC differential measured observations. On the one hand, this effectively suppresses estimation errors caused by sensor noise and external electromagnetic disturbances, and avoids path jumps caused by independent single-point inversion by utilizing the cable spatial continuity constraint. On the other hand, it can simultaneously output the observation residuals and posterior covariance, quantifying the reliability of the offset estimation results and providing an evaluation basis for subsequent multi-source feature fusion and cable ridge tracking.

[0055] As an optional implementation, step S4 includes: The triaxial characteristics of the high-frequency alternating magnetic field include: a triaxial amplitude vector and a phase vector; Step S41: Calculate the composite amplitude and its direction angle in the horizontal plane based on the three-axis amplitude vector. Based on the direction angle, obtain two candidate orientation angles for the current station, expressed as:

[0056]

[0057] Among them, A h For the synthesized amplitude, A x A represents the X-axis magnitude in the three-axis magnitude vector. y θ represents the Y-axis magnitude in the three-axis magnitude vector. h θ is the direction angle. c To move towards candidate angles; Step S42: Based on the two candidate orientation angles of the current station and the effective orientation of the previous station, determine the effective orientation candidate value of the current station, and traverse each station in turn to obtain the effective orientation candidate value of each station of the target buried cable. Step S43: Process the valid candidate values ​​of each station to obtain the trajectory of the target buried cable, wherein the trajectory includes the trajectory estimate of each station.

[0058] Specifically, the cable routing is locked using the triaxial characteristics of a high-frequency alternating magnetic field. At each station, the triaxial amplitude vector and phase vector reflect the directional distribution of the alternating magnetic field in space. When the detection equipment moves along the survey line, if there is an angle between the survey line and the target buried cable routing, the amplitude ratio of the triaxial amplitude vector and the phase difference of the phase vector will show a discernible pattern of change. Therefore, this invention obtains routing estimation based on the triaxial amplitude vector and phase vector, specifically: the composite amplitude and its direction angle are calculated based on the triaxial amplitude vector. A 90° shift in the direction angle from different directions yields two candidate routing angles, which are two candidate directions differing by 180°. The two candidate angles at the current station are compared with the determined routing at the previous station, and the candidate angle with the smaller angle difference is selected as the effective routing candidate value for that station, thus obtaining a unique preliminary routing estimate that conforms to path continuity. This process is repeated for each station to obtain the effective routing candidate value for each station. Next, the valid route candidate values ​​for each station are processed. First, unreliable station valid route candidate values ​​are eliminated. Then, a spatial distance-based smoothing process is applied to the remaining valid route candidate values. For example, a Kalman filter is used for smoothing. Interpolation is then performed on the eliminated points, ultimately outputting a continuous and smooth route trajectory. This process effectively suppresses random abrupt changes caused by local metals or heterogeneous media. At the same time, by detecting continuous changes in the route angle, it identifies and retains real cable bends or branch points, providing a priori basis for subsequent path tracing and branch node identification.

[0059] As an optional implementation method, please refer to the appendix. Figure 5 Step S5 includes: Step S51: Based on the lateral offset estimate and the orientation estimate, construct a positioning dataset, which includes: the station coordinates, lateral offset estimate, orientation estimate, DC component evaluation index, and high frequency component evaluation index for each station. Step S52: Using the spatial coordinates of each measurement station as a unified index, perform spatial serialization processing on the positioning dataset corresponding to each station to generate an ordered station sequence with coordinates increasing along the survey line. Step S53: Determine the center point of the survey line. Using the center point of the survey line as the spatial projection reference, perform spatial projection calculations by combining the lateral offset estimation and the direction estimation to obtain a set of candidate cable center points. Step S54: Read the DC component evaluation index and high frequency component evaluation index of each station sequentially along the ordered station sequence, and select the first station in the ordered station sequence whose DC component evaluation index and high frequency component evaluation index both meet the preset confidence threshold as the initial seed point. Step S55: Based on the initial seed point and the set of candidate cable center points, execute a combined strategy of ridge tracing and local path fitting to obtain the cable center path.

[0060] Specifically, evaluation indices for the DC component and high-frequency component are constructed for each station. The DC component evaluation indices include the posterior covariance and observation residual of the Kalman filter output in step S34. The high-frequency component evaluation indices include the signal-to-noise ratio and phase consistency of the three-axis characteristics of the high-frequency alternating magnetic field. A positioning dataset is constructed based on the lateral offset estimation, orientation estimation, station coordinates, DC component evaluation indices, and high-frequency component evaluation indices for each measurement station. Using the spatial coordinates of each measurement station as a unified index, the resulting positioning dataset for each station is spatially serialized to generate an ordered sequence of stations with coordinates increasing along the survey line. Due to the complex factors in the actual laying environment of underground cables, such as bends, slope changes, branch connections, and local metal interference, the location dataset may still have minor deviations in terms of spatial continuity, local anomalies, and consistency of burial depth estimation. To ensure that the output results meet the requirements of path continuity, accurate location, reliable burial depth, and verifiability and traceability in engineering applications, the cable center path is extracted based on the location dataset to facilitate subsequent positioning. Since the two sets of magnetometers are symmetrically arranged on the detection device, the geometric center of the line connecting the two sets of magnetometers is taken as the center point of the measurement line. The spatial projection calculation method is as follows: using the coordinates of the center point of the measurement line as a reference, the lateral offset projection orientation is determined by combining the cable direction estimation, and the lateral offset direction is perpendicular to the cable direction; based on the magnitude and sign of the estimated lateral offset, planar spatial projection is carried out along the lateral offset projection orientation to obtain the candidate coordinates of the cable center corresponding to the current station; the ordered station sequence is traversed to complete the calculation of all stations, and a set of candidate cable center points is generated. The positioning dataset in step S51 can summarize all valid information from all stations, including the positioning data of each station, namely: station coordinates, lateral offset estimation, orientation estimation, DC component evaluation index, and high-frequency component evaluation index. Since the order of stations collected during field exploration may be chaotic and the measurement points may be disordered, step S52 reorders the positioning data according to the direction of the survey line, generating an ordered station sequence. This allows subsequent ridgeline tracing to iterate sequentially. Because the lateral offset estimation is a scalar value, only representing the vertical distance between the center of the target buried cable and the center point of the survey line, it lacks spatial orientation information and cannot directly characterize the cable's geographical location. The orientation estimation, on the other hand, is the cable's extension azimuth angle in the geodetic coordinate system, representing the lateral offset direction of the cable relative to the center point of the survey line. Perpendicular to the cable route; therefore, spatial projection calculation is performed by combining lateral offset estimation and route estimation, and the lateral offset estimation is converted into the geodetic plane coordinates corresponding to the cable center. Batch calculation is completed by traversing all ordered sites, and a set of candidate cable center points is generated. The combination strategy of ridge tracing and local path fitting requires the selection of an initial starting point (initial seed point). In step S54, in order to avoid large noise in the initial seed point data and distortion of offset estimation, the initial seed point is selected based on the condition that both the DC component evaluation index and the high frequency component evaluation index meet the preset confidence threshold. In step S55, the combination strategy of ridge tracing and local path fitting is executed based on the initial seed point and the set of candidate cable center points to obtain a complete and continuous cable center path.

[0061] As an optional implementation, step S55 includes: Using the estimated direction from the initial seed point as the first-round tracking direction, adjacent stations are located; Select the cable center candidate points corresponding to adjacent stations from the set of cable center candidate points; Construct a spatial sliding window centered on the candidate point at the center of the cable and containing several neighboring candidate points; Within the sliding window, the least squares method is used to perform curve fitting on the set of all candidate points within the sliding window. The coordinates and tangent direction of the fitted curve at the current position are solved. Based on the solution results, the candidate points of the cable center are updated and corrected to obtain the corrected candidate points. The modified candidate points are determined as cable center ridge nodes and stored in the complete cable center path; The corrected candidate point is set as the initial seed point for the next round of tracing iteration, and the iteration continues to obtain the complete cable center path.

[0062] Specifically, based on the currently determined initial seed point The search direction is determined by using direction estimation, and the next adjacent station is searched in this direction. Extract the candidate cable center points corresponding to the adjacent sites from the set of candidate cable center points. ; based on the current candidate cable center point Centered on a central point, a spatial sliding window is constructed, encompassing several neighboring points. Since the local cable laying trajectory approximates a smooth curve, the true location cannot be determined solely by a single point. Therefore, the sliding window limits the local spatial range to suppress single-point anomalous noise. Within the window, the least squares method is used to perform curve fitting on the candidate point set within the window. The tangent direction and coordinates of the fitted curve at the current position are calculated, and this is used to update and replace the original candidate points. This is used as a definite ridge node to be included in the central path and as a seed point for the next round of tracking. This eliminates path reversals or jumps caused by single-point measurement errors through local smoothing. Through continuous iteration and correction, the next station is found by taking the previous optimized point as the benchmark. All measurement points are traversed point by point, and finally a continuous and complete cable central path is spliced ​​together.

[0063] As an optional implementation method, please refer to the appendix. Figure 6 Step S6 involves extracting the burial depth data of the target underground cable, including: Step S61: Construct a model of the relationship between magnetic field characteristics and burial depth based on the Biot-Savart law, the expression of which is:

[0064] in, For vertical burial depth estimation, The characteristic amplitude of the magnetic field. To inject current, For lateral offset estimation, Permeability of free space; Step S62: Obtain the amplitude of the DC stable magnetic field signal and the characteristic amplitude of the high-frequency magnetic field at the current site. Based on the amplitude of the DC stable magnetic field signal, the characteristic amplitude of the high-frequency magnetic field, and the relationship model, calculate the estimated value of the DC source burial depth and the estimated value of the high-frequency source burial depth. Step S63: Calculate the initial burial depth fusion value of the current site by weighted averaging the DC source burial depth estimate and the high-frequency source burial depth estimate. Step S64: Repeat steps S61-S63 to obtain the initial burial depth fusion values ​​for all stations; Step S65: Identify anomalies in the initial burial depth fusion values ​​of all sites, correct the identified initial burial depth fusion values, and obtain continuous cable burial depth data for the target underground cable.

[0065] Specifically, a relationship model between magnetic field characteristics and burial depth is constructed. The known injection current amplitude, estimated lateral offset, DC stable magnetic field signal amplitude, and high-frequency magnetic field characteristic amplitude (calculated based on the triaxial characteristics of the high-frequency alternating magnetic field) obtained in step S1 for the current site are substituted into the relationship model to calculate the estimated burial depth of the DC source. Compared with the estimated burial depth of the high-frequency source (That is: the estimated depth of the DC source is obtained based on the amplitude of the DC stable magnetic field signal, and the estimated depth of the high-frequency source is obtained based on the characteristic amplitude of the high-frequency magnetic field.) Next, a weighted average is calculated between the estimated burial depth of the DC source and the estimated burial depth of the high-frequency source to obtain the initial fused burial depth value for the site. The weights of the DC source burial depth estimate and the high-frequency source burial depth estimate are determined based on the DC component evaluation index and the high-frequency component evaluation index in the positioning dataset. Specifically, when the observation residual is small and the posterior covariance is low, the DC signal quality is considered excellent, and the signal is automatically improved. The weighting is adjusted to enhance the lateral constraint on the cable center position; conversely, when the triaxial characteristics of the high-frequency alternating magnetic field have a high signal-to-noise ratio and good phase consistency at the target frequency, the weighting is increased. The weights are adjusted to enhance the ability to lock onto the path direction; if a strong disturbance causes a certain mode to become less reliable, its weight is automatically reduced and compensated by another mode. Repeat steps S61-S63 to obtain the initial burial depth fusion values ​​for all stations; correct the identified abnormal initial burial depth fusion values ​​by setting a reasonable physical burial depth threshold (0.5m-5m) to remove extreme points that clearly violate engineering common sense; simultaneously calculate the first-order difference of the burial depth sequence, if the burial depth change between adjacent stations... If the gradient exceeds the preset threshold, the point is identified as a suspected anomalous jump point. For suspected anomalous jump points, a neighborhood weighted interpolation method is used to calculate the replacement value by weighting the burial depth values ​​of adjacent normal points in inverse proportion to the distance. For continuous abrupt change sections caused by changes in terrain slope, the abrupt change characteristics are preserved, and spline interpolation is used to smooth the transition section at the start and end of the abrupt change. Finally, a moving average filter is applied to the entire sequence to obtain continuous and smooth cable burial depth data along the cable center path. Based on the continuous cable burial depth data, a cable burial depth profile can be obtained to intuitively show the changes in cable burial depth along the line.

[0066] To ensure the verifiability and traceability of positioning results, this invention introduces an accuracy verification and error assessment mechanism to perform multi-dimensional consistency checks on cable path and burial depth outputs. Specifically, a comprehensive error index is constructed using observation residuals, posterior covariance, signal-to-noise ratio of high-frequency features, and phase consistency from the fusion process, forming a quality assessment of the positioning results for each station. If a segment exhibits persistently large observation residuals, significantly increased posterior covariance, or decreased phase consistency, it is determined that the segment has strong interference or decreased data reliability and is marked as a low-confidence segment. Simultaneously, the geometric continuity between adjacent stations is used as an auxiliary verification condition: if the centerline exhibits drastic curvature changes or unreasonable bends that do not conform to actual laying patterns locally, a path quality alarm is triggered, and the segment is marked as an abnormal segment. For scenarios with known reference points, the reference points can be further matched with the output path to calculate the deviation and use it for global or local correction, thereby improving the engineering consistency of the results.

[0067] Finally, the system outputs precise location information and reliability assessment results for the target buried cable. The output includes at least: a spatial coordinate sequence of the cable center path (the spatial coordinate sequence can be output by station or by equidistant sampling), path orientation information, lateral offset curves for each station, cable burial depth data along the path, and burial depth profiles. Specifically, the lateral offset and orientation estimates for all stations are spatially sequenced based on their spatial locations, integrating them to form path orientation information and lateral offset curves for each station. For identified turning points, suspected branch points, and strongly interfering sections, their locations and types are output using event tagging, facilitating subsequent on-site verification and engineering decisions. Simultaneously, based on reliability evaluation indicators, a reliability level or confidence interval is generated for each station or section, and high-reliability sections / low-reliability sections / abnormal sections are marked in the results, forming a visual representation that can directly serve construction avoidance and operation and maintenance diagnosis.

[0068] The above are merely specific embodiments of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. The scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A method for locating underground cables based on the fusion of composite electromagnetic fields and multi-source information, characterized in that, Includes the following steps: Step S1: Simultaneously inject a large ampere DC current and a high-frequency small current into the target buried cable using an AC-DC coupling current generator to form a magnetic field in the space around the target buried cable. Step S2: Use a multi-sensor collaborative acquisition device to acquire the DC stable magnetic field signal and the high-frequency alternating magnetic field signal of the magnetic field, and obtain the amplitude of the DC stable magnetic field signal and the triaxial characteristics of the high-frequency alternating magnetic field based on the DC stable magnetic field signal and the high-frequency alternating magnetic field signal, respectively. Step S3: Perform Kalman filtering analysis based on the amplitude of the DC stable magnetic field signal to obtain an estimate of the lateral offset of the target buried cable; Step S4: Estimate the direction of the target buried cable based on the triaxial characteristics of the high-frequency alternating magnetic field; Step S5: Based on the lateral offset estimate and the direction estimate, obtain the positioning dataset, and obtain the cable center path based on the positioning dataset; Step S6: Extract the burial depth data of the target underground cable, and output the positioning information of the target underground cable by combining the cable center path and the burial depth data.

2. The method for locating underground cables based on the fusion of composite electromagnetic fields and multi-source information according to claim 1, characterized in that, The AC-DC coupling current generating device includes: a high-power DC power supply, a frequency converter signal source, a coupler, a power amplifier, and an AC-DC coupling output module; The output terminal of the high-power DC power supply is electrically connected to the input terminal of the coupler, the frequency converter signal source is electrically connected to the input terminal of the coupler through the power amplifier, and the output terminal of the coupler is electrically connected to the input terminal of the AC-DC coupling output module.

3. The underground cable positioning method based on composite electromagnetic field and multi-source information fusion according to claim 1, characterized in that, The multi-sensor collaborative acquisition device includes: two sets of magnetometers, two sets of triaxial high-frequency coils, a follower, a frequency selector, an A / D converter, an amplifier, a CAN bus, and a CPU; The two sets of magnetometers are used to collect the DC stable magnetic field signal generated by the target buried cable. The signal output terminals of the two sets of magnetometers are connected to the signal input terminal of the CPU through the CAN bus. The amplitude of the DC stable magnetic field signal is obtained after processing by the CPU. The two sets of triaxial high-frequency coils are used to acquire the high-frequency alternating magnetic field signal generated by the target buried cable. The signal output terminals of the two sets of triaxial high-frequency coils are connected in series with the follower, the frequency selector, the amplifier, and the input terminal of the A / D converter. The output terminal of the A / D converter is connected to another signal input terminal of the CPU. The high-frequency alternating magnetic field triaxial characteristics are obtained after processing by the CPU.

4. The underground cable positioning method based on composite electromagnetic field and multi-source information fusion according to claim 1, characterized in that, Step S3 includes: Step S31: Set up stations along the survey line. Construct a state vector using the lateral offset of the target buried cable relative to the center of the survey line at each station and the rate of change of offset along the direction of the survey line station. The expression is: in, Let s be the state vector of station s. Let be the rate of change of offset at station s. Estimate the lateral offset at station s; Step S32: Based on the amplitude of the DC stable magnetic field signal from the two sets of magnetometers, obtain the DC differential measured observation, expressed as: in, For DC differential measured observations at station s, The amplitude of the stable DC magnetic field signal obtained by the electromagnetic meter located to the left of the target buried cable. The amplitude of the DC stable magnetic field signal obtained by the electromagnetic meter located on the right side of the target buried cable; Step S33: Construct a monotonic mapping function between the lateral offset and the DC magnetic field difference based on the Biot-Savart law, and construct the observation equation according to the monotonic mapping function, the expression of which is: in, Let s be the state vector of station s. Here is the state transition matrix. Let f be the state vector of the previous station. This is the monotonic mapping function between the DC magnetic field difference and the lateral offset, obtained according to the Biot-Savart law. To inject direct current, As the initial value for burial depth, The distance between the two sets of magnetometers is half-pitch. For noise in the offset change rate process, To observe Gaussian noise, The predicted observations at station s are obtained based on the lateral offset solution. Step S34: Traverse all measurement stations along the survey line in sequence, and perform Kalman filter prediction-update recursive estimation for each station in sequence to obtain the lateral offset estimate of the target buried cable.

5. The underground cable positioning method based on composite electromagnetic field and multi-source information fusion according to claim 4, characterized in that, Step S34 includes: The predicted state vector of the current station is obtained by predicting the current station using the posterior state vector and state transition matrix of the previous station. The expression is as follows: in, Predict the state vector for the current site. Let be the posterior state vector of the previous station. This is the state transition matrix; The prior prediction covariance of the current station is calculated using the posterior covariance of the previous station, expressed as: in, For the prior prediction covariance of the current site, For the posterior covariance of the previous site, For process noise covariance; The Kalman gain is calculated based on the prior prediction covariance of the current site, and the expression is as follows: in, For Kalman gain, For Jacobian matrices, To fix the observation noise covariance; Calculate the observation residual between the measured and predicted DC differential observations, and combine it with the Kalman gain to correct the current station's predicted state, thus obtaining the optimal posterior predicted state for the current station. The expression is as follows: in, This represents the optimal posterior state vector for the current site. For the predicted observations obtained from the prior predicted state, To observe the residuals, These are the DC differential measured observations at station s; The posterior covariance of the current site is updated synchronously using the following expression: in, The posterior covariance of the current site. It is an identity matrix.

6. The method for locating underground cables based on the fusion of composite electromagnetic fields and multi-source information according to claim 1, characterized in that, Step S4 includes: The triaxial characteristics of the high-frequency alternating magnetic field include: a triaxial amplitude vector and a phase vector; Step S41: Calculate the composite amplitude and its direction angle in the horizontal plane based on the three-axis amplitude vector. Based on the direction angle, obtain two candidate orientation angles for the current station, expressed as: Among them, A h For the synthesized amplitude, A x A represents the X-axis magnitude in the three-axis magnitude vector. y θ represents the Y-axis magnitude in the three-axis magnitude vector. h θ is the direction angle. c To move towards candidate angles; Step S42: Based on the two candidate orientation angles of the current station and the effective orientation of the previous station, determine the effective orientation candidate value of the current station, and traverse each station in turn to obtain the effective orientation candidate value of each station of the target buried cable. Step S43: Process the valid candidate values ​​of each station to obtain the trajectory of the target buried cable, wherein the trajectory includes the trajectory estimate of each station.

7. The method for locating underground cables based on the fusion of composite electromagnetic fields and multi-source information according to claim 1, characterized in that, Step S5 includes: Step S51: Based on the lateral offset estimate and the orientation estimate, construct a positioning dataset, which includes: the station coordinates, lateral offset estimate, orientation estimate, DC component evaluation index, and high frequency component evaluation index for each station. Step S52: Using the spatial coordinates of each measurement station as a unified index, perform spatial serialization processing on the positioning dataset corresponding to each station to generate an ordered station sequence with coordinates increasing along the survey line. Step S53: Determine the center point of the survey line. Using the center point of the survey line as the spatial projection reference, perform spatial projection calculations by combining the lateral offset estimation and the direction estimation to obtain a set of candidate cable center points. Step S54: Read the DC component evaluation index and high frequency component evaluation index of each station sequentially along the ordered station sequence, and select the first station in the ordered station sequence whose DC component evaluation index and high frequency component evaluation index both meet the preset confidence threshold as the initial seed point. Step S55: Based on the initial seed point and the set of candidate cable center points, execute a combined strategy of ridge tracing and local path fitting to obtain the cable center path.

8. The method for locating underground cables based on the fusion of composite electromagnetic fields and multi-source information according to claim 7, characterized in that, Step S55 includes: Using the estimated direction from the initial seed point as the first-round tracking direction, adjacent stations are located; Select the cable center candidate points corresponding to adjacent stations from the set of cable center candidate points; Construct a spatial sliding window centered on the candidate point at the center of the cable and containing several neighboring candidate points; Within the sliding window, the least squares method is used to fit a curve to the set of all candidate points within the sliding window. The coordinates and tangent direction of the fitted curve at the current position are solved. Based on the solution results, the original candidate points for the cable center are updated and corrected to obtain the corrected candidate points. The modified candidate points are determined as cable center ridge nodes and stored in the complete cable center path; The corrected candidate point is set as the initial seed point for the next round of tracing iteration, and the iteration continues to obtain the complete cable center path.

9. The method for locating underground cables based on the fusion of composite electromagnetic fields and multi-source information according to claim 7, characterized in that, The step S6 involves extracting the burial depth data of the target underground cable, including: Step S61: Construct a model of the relationship between magnetic field characteristics and burial depth based on the Biot-Savart law, the expression of which is: in, For vertical burial depth estimation, The characteristic amplitude of the magnetic field. To inject current, For lateral offset estimation, The vacuum permeability; Step S62: Obtain the amplitude of the DC stable magnetic field signal and the characteristic amplitude of the high-frequency magnetic field at the current site. Based on the amplitude of the DC stable magnetic field signal, the characteristic amplitude of the high-frequency magnetic field, and the relationship model, calculate the estimated value of the DC source burial depth and the estimated value of the high-frequency source burial depth. Step S63: Calculate the initial burial depth fusion value of the current site by weighted averaging the DC source burial depth estimate and the high-frequency source burial depth estimate. Step S64: Repeat steps S61-S63 to obtain the initial burial depth fusion values ​​for all stations; Step S65: Identify anomalies in the initial burial depth fusion values ​​of all sites, correct the identified initial burial depth fusion values, and obtain continuous cable burial depth data for the target underground cable.

Images