A method and system for measuring line current based on a multi-sensor array

Abstract

The application relates to the technical field of electric power measurement, and discloses a wire current measurement method and system based on a multi-sensor array. The method contacts the surface of a wire with an insulating probe with a positioning label to fit the axis of the wire to be measured; a distributed array composed of multiple integrated measurement chips is arranged near the wire, the number of the chips is greater than the number of the wires, and the chips are internally provided with single-axis magnetic field sensors, positioning and attitude units. The reference position and the spatial attitude of the chips are obtained, rigid body transformation compensation is performed on the basis of preset structure offset parameters and the attitude, actual measurement point positions and sensitive axis directions are obtained, geometric relationships are calculated according to the wire axis, the measurement point positions and the sensitive axis directions, and magnetic field response coefficients are determined; magnetic field projection observation values are collected, a multi-chip overdetermined observation system is constructed in combination with the response coefficients, and finally, a current estimation value is inversely solved by using a weighted least square algorithm. The application can realize high-precision current inversion in a multi-wire complex electromagnetic environment.

Description

Technical Field

[0001] This invention relates to the field of power measurement technology, specifically a method and system for measuring conductor current based on a multi-sensor array. Background Technology

[0002] The core principle of non-contact conductor current measurement is the physical mapping relationship between the magnetic field surrounding the current-carrying conductor and the conductor current. Ideally, by obtaining the magnetic induction intensity at a specific measurement point in space and combining this with the perpendicular distance from the measurement point to the conductor axis, the magnitude of the conductor current can be deduced. However, in practical engineering applications, existing measurement schemes face the following technical bottlenecks: (1) Existing technologies often use a single magnetic field sensor for measurement, based on the assumption that the sensor is located at a known distance and that the sensitive axis is perfectly aligned with the direction of the circular magnetic field of the conductor. However, in actual inspections or online monitoring, the relative distance and spatial angle between the sensor and the conductor under test are often random. Since a single-axis magnetic field sensor only responds to the projection of the magnetic field onto its sensitive axis, the output signal will be significantly attenuated when there is an angle between the sensitive axis and the direction of the spatial magnetic field. In this case, a single sensor cannot effectively distinguish whether the signal reduction is caused by a decrease in conductor current, an increase in measurement distance, or a spatial angle shift, which can easily lead to serious distortion of the inversion results.

[0003] (2) In real measurement sites, there are usually multiple interference components such as the geomagnetic field, the spatial magnetic field coupling of nearby current-carrying conductors, the residual magnetism of surrounding ferromagnetic equipment, and the zero-point drift of the sensor itself. Traditional single-sensor or simple direct substitution calculation mode cannot effectively remove these spatial interference quantities, resulting in weak anti-interference ability of the system.

[0004] (3) In application scenarios such as three-phase overhead lines, multi-circuit overhead lines on the same tower, or complex busbars, the number of conductors to be measured is often multiple, and the spatial magnetic field is a vector superposition of multiple unknown current sources. Although increasing the number of sensors can obtain multiple sets of observation data, in actual measurements, due to the inherent hardware noise, spatial positioning error, attitude deviation, and environmental interference of the sensors, the observation equations of each sensor usually cannot strictly converge to the same exact solution. Existing technologies often lack effective fusion mechanisms for multi-source heterogeneous errors, making it difficult to transform redundant spatial observation information with random errors into a highly robust global optimal solution, resulting in a significant reduction in measurement accuracy and engineering practicality in multi-conductor and strong interference environments.

[0005] In summary, existing technologies have significant shortcomings in handling spatial location uncertainties, complex magnetic field interference, and multi-sensor data fusion during non-contact flow measurement. Summary of the Invention

[0006] To address the technical problems existing in the prior art, this invention provides a method and system for measuring conductor current based on a multi-sensor array, aiming to effectively overcome the errors caused by the randomness of the spatial pose of the measuring equipment and achieve high-precision current inversion in complex electromagnetic environments with multiple conductors.

[0007] To achieve the above objectives, the present invention provides the following technical solution: This invention discloses a method for measuring conductor current based on a multi-sensor array, comprising the following steps: S1. Establish a global coordinate system and use an insulated calibration probe with a positioning label to contact the surface of the conductor under test to obtain spatial coordinate points, and then fit the axis of the conductor under test. S2. Deploy a distributed measurement array near the conductor to be tested; the distributed measurement array consists of multiple integrated measurement chips, and the number of integrated measurement chips is greater than the number of conductors to be tested for inverting the current; each integrated measurement chip is equipped with a magnetic field sensing unit, a positioning unit and an attitude sensing unit. S3. Obtain the reference point position and overall spatial attitude of each integrated measurement chip in the global coordinate system through the positioning unit and attitude sensing unit in each integrated measurement chip respectively; S4. Combining the preset structural offset parameters inside each integrated measurement chip with the overall spatial attitude, perform rigid body transformation compensation on the reference point position and the inherent sensitive axis direction inside the chip, and calculate the actual measurement point position and actual sensitive axis direction of each magnetic field sensing unit in the global coordinate system. S5. Based on the axis of the conductor to be tested, the actual measurement point position, and the actual sensitive axis direction, calculate the geometric and physical relationship between each actual measurement point and the conductor to be tested, and then determine the magnetic field response coefficient corresponding to each measurement point. S6. Obtain the magnetic field projection observation values ​​collected by each magnetic field sensing unit, and construct a multi-chip overdetermined observation system by combining the magnetic field response coefficients; S7. The weighted least squares algorithm is used to invert and solve the multi-chip overdetermined observation system to obtain the current estimate of the conductor to be measured.

[0008] As a further improvement to the above scheme, in step S4, the formulas for expressing the actual measurement point position and actual sensing axis direction of each magnetic field sensing unit in the global coordinate system are as follows: ; ; In the formula, the subscript For the index of the integrated measurement chip, , This refers to the actual number of integrated measurement chips installed and involved in the inversion process; For the first The actual measurement point location of each magnetic field sensing unit in the global coordinate system. For the first The reference point position obtained by each positioning unit; For the first The attitude rotation matrix obtained by the attitude sensor; The offset of the magnetic field sensing unit relative to the reference point in the coordinate system of the integrated measurement chip is a fixed structure. For the first The unit direction of each magnetic field sensing unit in the global coordinate system; This refers to the fixed sensitive axis direction of the magnetic field sensing unit in the chip coordinate system.

[0009] As a further improvement to the above scheme, step S5 specifically includes: Based on the axis of the conductor to be measured and the actual measuring point position, the radial vector perpendicularly pointing from the conductor to the actual measuring point is calculated. The magnitude of this vector is the vertical distance, expressed by the following formula: ; ; In the formula, the subscript For the index of the conductor to be tested, , The number of conductors to be tested for the current to be inverted; For from the first The axis of the conductor to be measured is perpendicular to the first... The shortest distance vector between the actual measurement points of each magnetic field sensing unit; For the first A point on the axis of the conductor to be measured; For the first The unit vector of the direction of the axis of the traverse to be measured; For the first The magnetic field sensing unit to the first The perpendicular distance from the axis of the conductor; According to the right-hand screw rule, the tangential unit direction of the magnetic field generated by the conductor at the actual measurement point is determined by the following formula: ; In the formula, For the first The test conductor is in the first... The tangential unit direction of the magnetic field generated at the actual measuring point of each magnetic field sensing unit; Based on the actual sensing axis direction, the magnetic field response coefficient of each magnetic field sensing unit to each conductor under test is calculated, expressed by the following formula: ; In the formula, For the first The magnetic field sensing unit for the first The magnetic field response coefficient of the conductor under test; is the vacuum permeability.

[0010] As a further improvement to the above scheme, the construction and solution process of the multi-chip overdetermined observation system specifically includes: By combining the observation equations of all integrated measurement chips, the system observation equations are constructed: ; In the formula, This is a vector composed of the observed values ​​projected onto the magnetic field. , For the first Magnetic field projection observations of each magnetic field sensing unit; superscript It is the transpose symbol; The response matrix is ​​a multi-chip non-square matrix composed of the response coefficients of each magnetic field. ; The vector of the current to be inverted, , For the first The current to be inverted in the conductor under test; This is the measurement error vector; The weighted least squares algorithm with a regularization term is used to solve the problem, and the inverse objective function is: ; In the formula, This is the current estimation vector after inversion; This is the weight matrix. , Represents a diagonal matrix. For the first The weighting coefficients corresponding to each integrated measurement chip; For regularization parameters; This is the regularization matrix; It is an L2 norm.

[0011] As a further improvement to the above scheme, the weighting coefficient The calculation formula is: ; In the formula, For the first The function of weighting coefficients corresponding to each integrated measurement chip. For the first Noise uncertainty of a single magnetic field sensing unit; For the first Positioning uncertainty of each positioning unit; For the first The attitude uncertainty of each attitude sensing unit; For the first The structural calibration residual error of the integrated measurement chip.

[0012] As a further improvement to the above scheme, step S7, after obtaining the estimated current value of the conductor to be tested, also includes the following steps: Based on the calculated current estimate and the measurement system equations, the inversion residuals corresponding to each integrated measurement chip are calculated using the following formula: ; In the formula, For the first The inversion residuals corresponding to each integrated measurement chip; For the first The estimated current value of the conductor to be tested; Determine whether each inversion residual exceeds the set residual threshold. When the inversion residual corresponding to a certain integrated measurement chip exceeds the threshold, assign a reduced weight coefficient to the chip, or remove it from the multi-chip overdetermined observation system and re-execute the weighted least squares algorithm for inversion calculation.

[0013] As a further improvement to the above scheme, when the current in the conductor to be measured is alternating current, the conductor current measurement method further includes: By repeatedly performing the inversion solution at multiple consecutive sampling times, the instantaneous current sequence of each conductor under test is obtained: ; In the formula, For each conductor to be tested on the 1st The instantaneous current sequence at each sampling time; For the first The test conductor is in the first... The estimated instantaneous current at each sampling time; Based on the instantaneous current sequence, the AC characteristic parameters of each conductor under test are calculated, and the AC characteristic parameters include at least one of effective value, peak value and phase.

[0014] As a further improvement to the above scheme, in step S2, the following evaluation function is used to select the optimal chip placement combination for actual observation from several integrated measurement chips: ; In the formula, The optimal chip placement combination; This refers to one type of chip arrangement combination; Chip placement and assembly The corresponding response matrix condition number; Indicates the arrangement of chips The inversion error is predicted by a variety of factors. Chip placement and assembly The corresponding deployment cost; These are the weighting coefficients.

[0015] As a further improvement to the above scheme, in step S1, the insulation calibration probe with the positioning tag sequentially contacts multiple positions on the surface of the conductor to be tested and records the corresponding spatial contact points; when there is a fixed offset between the contact tip of the insulation calibration probe and the positioning tag, the recorded contact points are corrected in combination with the fixed offset to calculate the accurate actual contact point position.

[0016] This invention also discloses a conductor current measurement system based on a multi-sensor array, used to implement the conductor current measurement method based on a multi-sensor array as described above; the conductor current measurement system includes: Multiple positioning reference nodes are deployed around the measurement area and used to establish the global coordinate system; An insulating calibration probe with a positioning label attached to its surface is used to contact the surface of the conductor under test to generate coordinate points for fitting. Insulation measuring bracket, including several mounting positions; The distributed measurement array consists of multiple integrated measurement chips, with the number of integrated measurement chips exceeding the number of conductors whose current is to be inverted. Each integrated measurement chip contains a magnetic field sensing unit, a positioning unit, and an attitude sensing unit. Multiple integrated measurement chips are distributed and installed in at least one mounting grid on an insulated measurement bracket, with positional differences in the lateral, height, and depth directions. The optimal combination of chips for actual observation is selected from multiple candidate mounting grids on the insulated measurement bracket based on an evaluation function composed of the response matrix condition number and prediction error.

[0017] Compared with the prior art, the beneficial effects of the present invention are: 1. This invention discloses a multi-sensor array-based method for measuring conductor current. By integrating a positioning unit, an attitude sensing unit, and a single-axis magnetic field sensing array, and combining a weighted least squares overdetermined inversion algorithm, it achieves high-accuracy non-contact measurement of multi-conductor current under complex spatial poses. This invention utilizes the positioning unit and attitude sensing unit to obtain the spatial coordinates and sensitive axis orientation of each measurement point, and calculates the corresponding magnetic field projection coefficients. In multi-conductor scenarios, when the number of measurement observations exceeds the number of currents to be inverted, multiple measurement constraints with errors are transformed into a weighted residual minimization problem. Spatial pose compensation reduces the error dispersion of single-point measurements, fully utilizes redundant measurement information, and improves the accuracy and stability of optimal current estimation.

[0018] 2. This invention achieves background interference separation in complex electromagnetic environments by constructing a joint observation matrix mechanism that includes the measured current and three-dimensional background magnetic field components. The system uses multiple sensing units with different spatial positions or sensitive axis directions for joint solution, incorporating both the conductor current and the three-dimensional components of the background magnetic field into the inversion model, effectively eliminating the influence of unknown environmental magnetic fields or equipment residual magnetism on the target current measurement.

[0019] 3. This invention achieves highly robust inversion performance of the measurement system under local interference by employing an anomaly point identification and dynamic adjustment strategy based on inversion residuals. When the inversion residual of a certain measurement chip is determined to exceed a threshold or its response coefficient is too low, it is either weighted down or removed from the overdetermined observation matrix and then re-solved. This strategy avoids distortion of the overall current estimation caused by a single anomaly node, effectively enhancing the anti-interference capability of multi-sensor fusion.

[0020] 4. This invention achieves effective extraction of AC current characteristic parameters through a strategy of synchronous sampling of alternating magnetic fields and instantaneous sequence solving. The collected alternating magnetic field projection signal is calculated synchronously, and based on the inverted instantaneous current sequence, waveform parameters such as effective value, peak value, or phase are further calculated, meeting the engineering application requirements of practical power distribution monitoring. Attached Figure Description

[0021] Figure 1 This is a flowchart of the conductor current measurement method based on a multi-sensor array in Embodiment 1 of the present invention.

[0022] Figure 2 This is a schematic diagram of the layout of the conductor current measurement system based on a multi-sensor array in Embodiment 1 of the present invention.

[0023] Figure 3 This is a schematic diagram of the radial vector pointing perpendicularly from the conductor to be measured to the actual measuring point in Embodiment 1 of the present invention.

[0024] Figure 4 This is a schematic diagram illustrating the relationship between the multi-measurement chip and the multi-conductor space observation in Embodiment 1 of the present invention.

[0025] Figure 5 This is a comparison chart of current inversion errors under different observation redundancy levels in Embodiment 1 of the present invention.

[0026] Figure 6 This is a comparison chart of the direct solution and the overdetermined optimal estimation results in Embodiment 1 of the present invention.

[0027] Figure 7 This is a box plot showing the error of different current inversion methods under random pose conditions in Embodiment 1 of the present invention.

[0028] Figure 8 This is a comparison chart of the average errors of different current inversion methods in Embodiment 1 of the present invention.

[0029] Figure 9 This is a graph showing the change in magnetic field strength around the conductor as a function of distance in Embodiment 1 of the present invention.

[0030] Figure 10 This is a diagram showing the influence of attitude angle on the measurement component in the TMR single-axis sensitivity characteristic of Embodiment 1 of the present invention. Detailed Implementation

[0031] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0032] This embodiment provides a conductor current measurement system based on a multi-sensor array, including: a positioning reference node, an insulation calibration probe, an insulation measurement bracket, and a distributed measurement array.

[0033] Multiple positioning reference nodes are set up and deployed around the measurement area to establish a global coordinate system. In this embodiment, the positioning reference nodes can be UWB base stations, and four or more are set up.

[0034] The insulation calibration probe has a UWB positioning tag attached to its surface and is used to contact the surface of the conductor under test to generate coordinate points for fitting.

[0035] The distributed measurement array consists of multiple integrated measurement chips, and the number of integrated measurement chips is greater than the number of conductors to be measured for inverting the current; each integrated measurement chip is equipped with a magnetic field sensing unit, a positioning unit, and an attitude sensing unit.

[0036] In this embodiment, the magnetic field sensing unit can be a tunneling magnetoresistive (TMR) sensor, whose internal structure includes: a magnetic layer with a fixed magnetization direction (the reference direction remains unchanged), an extremely thin insulating barrier layer (electrons pass through via quantum tunneling), and a free magnetization layer (its direction rotates with the external magnetic field). When an external magnetic field is present, the magnetization direction of the free magnetization layer deflects, the electron tunneling probability changes, resulting in a significant change in the overall resistance, which is then converted into a differential voltage output. In some embodiments, the magnetic field sensing unit can also be of other types, such as a Hall sensor, an anisotropic magnetoresistive (AMR) sensor, a giant magnetoresistive (GMR) sensor, or a fluxgate sensor.

[0037] The positioning unit can use a UWB tag. The UWB tag and the UWB base station are located by ranging / time difference of arrival, which gives the position of the measurement chip in the global coordinate system.

[0038] The attitude sensing unit can be an IMU (Inertial Measurement Unit). One IMU contains three single-axis accelerometers and three single-axis gyroscopes. The accelerometers detect the acceleration signals of the object in the independent three axes of the carrier coordinate system, while the gyroscopes detect the angular velocity signals of the carrier relative to the navigation coordinate system. The angular velocity and acceleration of the object in three-dimensional space are measured, and the attitude of the object is calculated accordingly.

[0039] The magnetic field sensing unit, positioning unit, and attitude sensing unit are integrated onto a single measurement chip or rigid measurement substrate, allowing for unified calibration of position and attitude.

[0040] The insulation measurement bracket includes several mounting slots. Multiple integrated measurement chips are distributed and mounted in at least one mounting slot on the insulation measurement bracket, with positional differences in the lateral, vertical, and depth directions. The object being measured can be a long straight conductor, ensuring that the projections of each magnetic field sensing unit onto the plane of the conductor's cross-section do not overlap. An evaluation function based on the response matrix condition number and prediction error is used to select the optimal chip placement combination for actual observation from multiple candidate mounting slots on the insulation measurement bracket. In some embodiments, the bottom of the insulation measurement bracket may be equipped with casters and / or a telescopic structure, allowing for flexible adjustment of position, height, or angle to adapt to different deployment environments.

[0041] Please see Figure 1 This embodiment also provides a method for measuring conductor current based on a multi-sensor array. Using the above-mentioned measurement system, the method includes the following steps, namely S1 to S7.

[0042] S1. Establish a global coordinate system and use an insulated calibration probe with a positioning label to contact the surface of the conductor under test to obtain spatial coordinate points, and then fit the axis of the conductor under test.

[0043] In step S1, the insulation calibration probe with the positioning tag sequentially contacts multiple positions on the surface of the conductor to be tested and records the corresponding spatial contact points; when there is a fixed offset between the contact tip of the insulation calibration probe and the positioning tag, the recorded contact points are corrected in combination with the fixed offset to calculate the accurate actual contact point position.

[0044] First, multiple external UWB positioning reference nodes are deployed around the measurement area, and a global coordinate system is established. The conductor under test does not emit a UWB signal, therefore its position cannot be directly obtained from UWB. The conductor axis is obtained using an insulation calibration probe with a UWB tag.

[0045] During calibration, the insulating probe sequentially contacts multiple locations on the surface of the conductor under test to obtain the coordinates of the contact points: ; For a conductor that is approximately a straight line, the first... Root conductor axis: ; in, Let be a point on the axis of the conductor. Let be the unit vector along the axis of the conductor. These are the parameters for a straight line.

[0046] If two contact points are used and Simplified calibration can be adopted as follows: ; For the first The unit vector of the direction of the axis of the traverse to be measured is obtained. At this point, the spatial position and orientation of the traverse to be measured in the global coordinate system can be determined.

[0047] S2. Deploy a distributed measurement array near the conductor to be tested; the distributed measurement array consists of multiple integrated measurement chips, and the number of integrated measurement chips is greater than the number of conductors to be tested for inverting the current; each integrated measurement chip is equipped with a magnetic field sensing unit, a positioning unit and an attitude sensing unit.

[0048] Please see Figure 2 , Figure 1 In the diagram, 1 represents the conductor under test, 2 represents the insulation measurement bracket, 3 represents the mounting grid, and 4 represents the chip under test. There are three conductors under test in total, with three perforated insulation measurement brackets placed beneath them. When deploying the measurement system, the plane of each insulation measurement bracket should be kept as parallel as possible to the axis of the conductor under test to ensure that an integrated measurement chip is present below the projection of each conductor on the horizontal plane. The number of integrated measurement chips on each insulation measurement bracket varies, and their positions differ in three dimensions.

[0049] When there are multiple conductors, each magnetic field sensing unit measures the superposition of the magnetic fields of multiple conductors. If a fixed arrangement is not set so that all chips are very close to one conductor and far from the others, some columns in the response matrix will be weak, making it difficult to decouple the currents of some conductors. Therefore, "having a chip near the projection of each conductor" also serves another purpose: to ensure that each conductor has sufficiently effective observation constraints in the response matrix, improving decoupling accuracy. Whether directly below or diagonally below, the main goal is to have a measurement chip near the projection of each conductor. This ensures that each conductor has sufficient observation constraints, and the columns of the response matrix K have discriminative power, guaranteeing that weighted least squares inversion can accurately decouple the currents of multiple conductors.

[0050] In addition, considering engineering implementation, the chip can be fixed in the preset hole position by combining a vertical hole array or grid bracket, and then the bracket can be moved as a whole to below or diagonally below the wire to achieve a layout that is both fixed and adjustable.

[0051] For an adjustable aperture array insulation measurement bracket, let the set of all its candidate mounting positions be... for: ; in, This represents the total number of candidate mounting holes. Indicates the first There are 10 candidate well locations. Candidate well locations can come from a single vertical well array measuring plate or from a combination of multiple measuring plates.

[0052] The actual chip combination involved in the inversion is: ; in, This refers to the actual number of integrated measurement chips installed and participating in the inversion. Since each integrated measurement chip in this invention contains only one magnetic field sensing unit, each selected mounting hole corresponds to one integrated measurement chip, one TMR measurement point, and one magnetic field observation equation.

[0053] In this embodiment, the following evaluation function can be used to select the optimal chip placement combination for actual observation from several integrated measurement chips: ; In the formula, The optimal chip placement combination; This refers to one type of chip arrangement combination; Chip placement and assembly The corresponding response matrix condition number; Indicates the arrangement of chips The inversion error is predicted by a variety of factors. Chip placement and assembly The corresponding deployment cost; These are the weighting coefficients.

[0054] S3. Obtain the reference point position and overall spatial attitude of each integrated measurement chip in the global coordinate system through the positioning unit and attitude sensing unit in each integrated measurement chip.

[0055] S4. Combining the preset structural offset parameters inside each integrated measurement chip with the overall spatial attitude, perform rigid body transformation compensation on the reference point position and the inherent sensitive axis direction inside the chip to calculate the actual measurement point position and actual sensitive axis direction of each magnetic field sensing unit in the global coordinate system.

[0056] In this embodiment, each integrated measurement chip has only one single-axis TMR. The UWB directly measures the position of the UWB positioning reference point in the chip, rather than the actual measurement point position of the TMR; the IMU measures the overall attitude of the chip, rather than the error-free direction of the TMR sensitive axis.

[0057] Since the single-axis TMR, UWB positioning reference point, and IMU are fixedly mounted on the same rigid chip substrate, their relative positions and orientations can be predetermined through structural calibration. Therefore, compensation is required through rigid body coordinate transformation.

[0058] In step S4, the formulas for expressing the actual measurement point position and actual sensing axis direction of each magnetic field sensing unit in the global coordinate system are as follows: ; ; In the formula, the subscript For the index of the integrated measurement chip, , This refers to the actual number of integrated measurement chips installed and involved in the inversion process; For the first The actual measurement point location of each magnetic field sensing unit in the global coordinate system. For the first The reference point position obtained by each positioning unit; For the first The attitude rotation matrix obtained by the attitude sensor; The offset of the magnetic field sensing unit relative to the reference point in the coordinate system of the integrated measurement chip is a fixed structure. For the first The unit direction of each magnetic field sensing unit in the global coordinate system; This refers to the fixed sensitive axis direction of the magnetic field sensing unit in the chip coordinate system.

[0059] Therefore, the method of this invention does not directly equate the UWB position with the TMR position, nor does it directly equate the original IMU angle with the TMR sensitive axis direction. Instead, it compensates based on the fixed structural offset and attitude matrix. The UWB positioning error, IMU attitude error, and structural calibration residual error can then be used for weight matrix or residual judgment.

[0060] S5. Based on the axis of the conductor to be tested, the actual measurement point position, and the actual sensitive axis direction, calculate the geometric and physical relationship between each actual measurement point and the conductor to be tested, and then determine the magnetic field response coefficient corresponding to each measurement point.

[0061] Step S5 specifically includes: Based on the axis of the conductor to be measured and the actual measuring point position, the radial vector perpendicularly pointing from the conductor to the actual measuring point is calculated. The magnitude of this vector is the vertical distance, expressed by the following formula: ; In the formula, the subscript For the index of the conductor to be tested, , The number of conductors to be tested for the current to be inverted; For from the first The axis of the conductor to be measured is perpendicular to the first... The shortest distance vector between the actual measurement points of each magnetic field sensing unit; in other words... The description is the first The TMR measurement point relative to the first The transverse position of the conductor axis. Its direction is the radial direction of the conductor cross-section, and its modulus is the perpendicular distance from the TMR measuring point to the conductor axis.

[0062] The radial quantity is defined in this invention because, for an approximately long straight conductor, the magnitude of the surrounding magnetic field is related to the perpendicular distance from the measuring point to the conductor's axis. The magnitude of the magnetic field can be expressed as: ; in, For conductor current, The permeability of free space, This is the perpendicular distance from the measuring point to the conductor axis.

[0063] Therefore, before calculating the magnetic field strength at the TMR measuring point, it is necessary to first determine the shortest perpendicular distance from that measuring point to the conductor axis. The vector corresponding to this shortest distance is the radial quantity. Its length is: The diagram is as follows Figure 3 As shown.

[0064] According to the right-hand screw rule, the tangential unit direction of the magnetic field generated by the conductor at the actual measurement point is determined by the following formula: ; In the formula, For the first The test conductor is in the first... The tangential unit direction of the magnetic field generated at the actual measurement point of each magnetic field sensing unit.

[0065] No. Root conductor current In the The magnetic induction intensity generated at each TMR measurement point is: ; In the formula, For the first The current in the conductor to be tested.

[0066] Since the TMR is a uniaxial magnetic field sensor, its output corresponds to the projection of the magnetic field along the sensitive axis. Therefore, the first... The TMR pair of the first The response coefficient of the conductor current is: ; In the formula, For the first The magnetic field sensing unit for the first The magnetic field response coefficient of the conductor under test.

[0067] S6. Obtain the magnetic field projection observation values ​​collected by each magnetic field sensing unit, and construct a multi-chip overdetermined observation system by combining the magnetic field response coefficient.

[0068] To more closely approximate the actual TMR measurement process, the measurement link can be divided into four levels: "spatial magnetic field - sensitive axis projection - TMR voltage output - acquisition noise".

[0069] No. The magnetic field projection value of a TMR can be expressed as: ; In the formula, For the first Measurement error of each magnetic field sensing unit.

[0070] If the background magnetic field is considered Then the observation equation can be written as: ; The TMR voltage output can be expressed by the following formula: ; In the formula, For the first The output voltage of each magnetic field sensing unit For the first Zero-point bias of each magnetic field sensing unit For the first The sensitivity coefficient of each magnetic field sensing unit For the first Voltage noise of each magnetic field sensing unit. Amplifier gain should be considered. and ADC midpoint voltage Then the ADC input voltage It can be represented as: ; In the formula, For the first The voltage noise or quantization error corresponding to each magnetic field sensing unit after ADC quantization is, i.e.

[0071] When calculating the magnetic field projection from the acquired voltage, it can be written as: ; In the formula, For the first The estimated magnetic field projection value of each magnetic field sensing unit is obtained by back-calculation from the ADC output voltage, taking into account zero-point bias, sensitivity, background magnetic field, random noise and ADC quantization error.

[0072] This model illustrates that the actual input inversion algorithm... It is not an ideal magnetic field value, but a measurement result that includes zero bias, sensitivity error, background magnetic field, random noise, amplifier error and ADC quantization error.

[0073] S7. The weighted least squares algorithm is used to invert and solve the multi-chip overdetermined observation system to obtain the current estimate of the conductor to be measured.

[0074] The construction and solution process of the multi-chip overdetermined observation system specifically includes: By combining the observation equations of all integrated measurement chips, the system observation equations are constructed: ; In the formula, This is a vector composed of the observed values ​​projected onto the magnetic field. , For the first Magnetic field projection observations of each magnetic field sensing unit; superscript It is the transpose symbol; The response matrix is ​​a multi-chip non-square matrix composed of the response coefficients of each magnetic field. ; The vector of the current to be inverted. , For the first The current to be inverted in the conductor under test; This is the measurement error vector.

[0075] The present invention measures the number of chips. Greater than the number of currents to be inverted ,but for The non-square matrix response matrix leads to overdetermined observations. In this case, only one should be selected. Even when solving directly for a single chip, redundant observations should not be discarded; instead, all valid observation equations should be incorporated into the optimal estimation process.

[0076] Due to the presence of UWB positioning errors, IMU attitude errors, TMR noise, and structural calibration residual errors in actual measurements, the various observation equations typically cannot strictly satisfy the same exact solution. Therefore, this method transforms current inversion into a weighted residual minimization problem: ; Its analytical solution is: ; Among them, the weight matrix for: ; In the formula, Represents a diagonal matrix; For the first The weighting coefficients corresponding to each integrated measurement chip can be determined by the measurement reliability, and the calculation formula is as follows: ; In the formula, For the first The function of weighting coefficients corresponding to each integrated measurement chip. For the first Noise uncertainty of a single magnetic field sensing unit; For the first Positioning uncertainty of each positioning unit; For the first The attitude uncertainty of each attitude sensing unit; For the first The structural calibration residual error of the integrated measurement chip.

[0077] When the response matrix is ​​ill-conditioned or noisy, a regularization term can be introduced, and the inversion objective function becomes: ; In the formula, This is the current estimation vector after inversion; For regularization parameters; This is the regularization matrix; It is an L2 norm.

[0078] Its analytical solution is: ; The core significance of this stage is that redundant observations obtained from multiple integrated measurement chips will not be discarded, but will instead be used as weak constraints in the optimal current estimation under the meaning of minimum residual.

[0079] After obtaining the estimated current value of the conductor under test, the residual vector is calculated: ; No. The residual of each measurement chip is: ; In the formula, For the first The inversion residuals corresponding to each integrated measurement chip; For the first The estimated current value of the conductor to be tested; Determine whether each inversion residual exceeds a set residual threshold. When the inversion residual corresponding to a certain integrated measurement chip exceeds the threshold, it can be determined that the first residual is... The chip may suffer from issues such as localized magnetic interference, unfavorable attitude, abnormal UWB positioning, abnormal IMU attitude, TMR malfunction, or insufficient signal-to-noise ratio due to excessive distance. A reduced weighting coefficient is assigned to this chip, expressed by the weighting formula: ; Alternatively, the chip can be removed from the multi-chip overdetermined observation system, and the weighted least squares algorithm can be re-executed for inversion calculation.

[0080] The least squares objective function mentioned above is actually based on the entire current vector. Solving for the unknowns, therefore the result is... It is also It is a dimensional vector, not a single scalar. Expanding it, it can be written as: ; in, They represent the 1st root, the 2nd root, ..., the 3rd root respectively. The current estimate of the conductor under test.

[0081] For alternating current, the instantaneous current sequence of each conductor under test can be obtained by repeatedly performing the inversion solution at multiple consecutive sampling times: ; In the formula, For each conductor to be tested on the 1st The instantaneous current sequence at each sampling time; For the first The test conductor is in the first... The estimated instantaneous current at each sampling time; Based on the instantaneous current sequence, AC characteristic parameters of each conductor under test are calculated, including at least one of RMS value, peak value, and phase. For example, the first... Effective value of the conductor under test for: ; In the formula, To set a cycle.

[0082] To verify the effectiveness of the proposed method, this embodiment also establishes a numerical simulation example of non-contact current measurement of three-phase conductors. Assume the number of currents to be inverted is m=3, and multiple integrated measurement chips are deployed near the conductor to be measured. Each measurement chip provides a magnetic field projection observation, and the corresponding response coefficient is calculated based on its spatial position and the direction of the sensitive axis. This simulation is not intended to demonstrate that a larger number of measurement chips necessarily leads to better performance, but rather to illustrate that when redundant observations (n>m) have been obtained, the proposed weighted least squares or regularized weighted least squares method can incorporate redundant observations as weak constraints into the overall residual minimization process, avoiding the discarding of known measurement information.

[0083] like Figure 4 As shown, multiple measurement chips are distributed near multiple conductors under test. The TMR sensing axes of each chip are in different directions, and the magnetic field projection values ​​measured by them together constitute the observation vector. Since the distance and sensing axis projection angle of each measurement chip relative to each conductor are different, their corresponding response coefficients are also different. Therefore, multiple measurement chips can jointly form an n-row m-column response matrix.

[0084] like Figure 5 As shown, the horizontal axis represents the number of measurement chips n involved in the inversion, and the number of currents to be inverted is fixed at m=3. When n=m, the number of observation equations is the same as the number of unknown currents, and measurement noise, positioning error, and attitude error will directly affect the solution results. When n>m, the system forms an overdetermined observation model, and multiple observation equations with errors are fused together by minimizing the weighted residuals. This result shows that redundant observation information can be transformed into optimal current estimation constraints in the sense of minimum residuals, rather than being discarded or simply averaged.

[0085] like Figure 6As shown, when directly solving using only m measurement chips, the response matrix is ​​close to a square matrix, making it easier to amplify individual measurement errors. When using n>m measurement chips and performing overdetermined optimal estimation, multiple sensor observations participate in the inversion as redundant weak constraints, which can reduce the impact of a single abnormal measurement point or random error on the final current estimate. The above simulation results provide an exemplary verification of the multi-sensor overdetermined fusion inversion method of the present invention, but do not limit the number of measurement chips, the layout, or the specific application scenarios of the present invention.

[0086] To further illustrate the relationship between the above-mentioned overdetermined redundant observation inversion approach and the actual TMR measurement link, this embodiment can also provide comparative simulation results based on inversion comparison, showing the results of traditional single-point direct inversion, distance compensation, distance-attitude joint compensation, and multi-measurement point fusion inversion. The following figures are only simulation verification examples, used to explain the necessity of UWB distance compensation, IMU attitude compensation, and multi-TMR measurement point fusion inversion in this invention.

[0087] like Figure 7 As shown, under random distance and random attitude angle conditions, the error distribution range of the traditional single-point direct inversion method is relatively wide, indicating that it is more sensitive to the measurement distance and sensitive axis direction. After introducing UWB distance compensation and IMU attitude compensation, the error distribution gradually converges. The multi-measurement point fusion inversion method can further reduce the error dispersion by integrating the observations of multiple TMR measurement points through weighted least squares, which reflects the suppression effect of multi-observation constraints on random errors.

[0088] like Figure 8 As shown, the average relative errors of the four methods differ significantly. Traditional single-point direct inversion has a large average error because it does not consider distance and attitude angle changes; distance compensation and distance-attitude joint compensation can gradually improve the inversion results; multi-point fusion inversion obtains an estimate with a smaller overall residual based on the integration of multiple TMR measurements, thus further reducing the average error.

[0089] like Figure 9 As shown, under the condition of a constant conductor current, the magnetic field strength at the measuring point decreases as the distance between the sensor and the conductor increases. This result indicates that non-contact current measurement is significantly dependent on spatial position; if the distance between the measuring chip and the conductor is not estimated and compensated, changes in distance can easily be mistaken for changes in current. Therefore, this invention obtains the spatial position through a UWB positioning unit and incorporates the distance factor into the calculation of the inversion coefficients.

[0090] like Figure 10As shown, the TMR single-axis sensor can only measure the projected component of the magnetic field along its sensitive axis. As the angle between the TMR sensitive axis and the magnetic field direction increases, the measured component gradually decreases; when the angle approaches 90 degrees, the output component approaches zero. This result indicates that attitude angle error significantly affects single-point measurement results. Therefore, this invention uses IMU attitude calculation to determine the sensitive axis direction and introduces attitude projection compensation into the inversion model.

[0091] In conclusion, Figures 7 to 10 The simulation verification approach of this invention is further explained from three aspects: error statistics, distance influence, and attitude angle influence. The technical route of this invention, which combines UWB positioning compensation, IMU attitude compensation, and multi-TMR measurement point fusion inversion, can reduce the impact of single-point measurement error on the current inversion results through geometric compensation and multi-measurement point fusion under the conditions of distance variation, attitude variation, and measurement noise.

[0092] The above description is only a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and inventive concept of the present invention, should be covered within the scope of protection of the present invention.

Claims

1. A method for measuring conductor current based on a multi-sensor array, characterized in that, Includes the following steps: S1. Establish a global coordinate system and use an insulated calibration probe with a positioning label to contact the surface of the conductor under test to obtain spatial coordinate points, and then fit the axis of the conductor under test. S2. Deploy a distributed measurement array near the conductor to be tested; the distributed measurement array consists of multiple integrated measurement chips, and the number of integrated measurement chips is greater than the number of conductors to be tested for inverting the current; each integrated measurement chip is equipped with a magnetic field sensing unit, a positioning unit and an attitude sensing unit. S3. Obtain the reference point position and overall spatial attitude of each integrated measurement chip in the global coordinate system through the positioning unit and attitude sensing unit in each integrated measurement chip respectively; S4. Combining the preset structural offset parameters within each integrated measurement chip with the overall spatial attitude, perform rigid body transformation compensation on the reference point position and the inherent sensitive axis direction within the chip to calculate the actual measurement point position and actual sensitive axis direction of each magnetic field sensing unit in the global coordinate system; wherein, the expression formulas for the actual measurement point position and actual sensitive axis direction of each magnetic field sensing unit in the global coordinate system are as follows: In the formula, the subscript For the index of the integrated measurement chip, , This refers to the actual number of integrated measurement chips installed and involved in the inversion process; For the first The actual measurement point location of each magnetic field sensing unit in the global coordinate system. For the first The reference point position obtained by each positioning unit; For the first The attitude rotation matrix obtained by the attitude sensor; The offset of the magnetic field sensing unit relative to the reference point in the coordinate system of the integrated measurement chip is a fixed structure. For the first The unit direction of each magnetic field sensing unit in the global coordinate system; The direction of the fixed sensing axis of the magnetic field sensing unit in the chip coordinate system; S5. Based on the axis of the conductor to be tested, the actual measurement point position, and the actual sensitive axis direction, calculate the geometric and physical relationship between each actual measurement point and the conductor to be tested, and then determine the magnetic field response coefficient corresponding to each measurement point. S6. Obtain the magnetic field projection observation values ​​collected by each magnetic field sensing unit, and construct a multi-chip overdetermined observation system by combining the magnetic field response coefficients; S7. The weighted least squares algorithm is used to invert and solve the multi-chip overdetermined observation system to obtain the current estimate of the conductor to be measured.

2. The method for measuring conductor current based on a multi-sensor array according to claim 1, characterized in that, Step S5 specifically includes: Based on the axis of the conductor to be measured and the actual measuring point position, the radial vector perpendicularly pointing from the conductor to the actual measuring point is calculated. The magnitude of this vector is the vertical distance, expressed by the following formula: In the formula, the subscript For the index of the conductor to be tested, , The number of conductors to be tested for the current to be inverted; For from the first The axis of the conductor to be measured is perpendicular to the first... The shortest distance vector between the actual measurement points of each magnetic field sensing unit; For the first A point on the axis of the conductor to be measured; For the first The unit vector of the direction of the axis of the traverse to be measured; For the first The magnetic field sensing unit to the first The perpendicular distance from the axis of the conductor; According to the right-hand screw rule, the tangential unit direction of the magnetic field generated by the conductor at the actual measurement point is determined by the following formula: In the formula, For the first The test conductor is in the first... The tangential unit direction of the magnetic field generated at the actual measuring point of each magnetic field sensing unit; Based on the actual sensing axis direction, the magnetic field response coefficient of each magnetic field sensing unit to each conductor under test is calculated, expressed by the following formula: In the formula, For the first The magnetic field sensing unit for the first The magnetic field response coefficient of the conductor under test; is the vacuum permeability.

3. The method for measuring conductor current based on a multi-sensor array according to claim 2, characterized in that, The construction and solution process of the multi-chip overdetermined observation system specifically includes: By combining the observation equations of all integrated measurement chips, the system observation equations are constructed: In the formula, This is a vector composed of the observed values ​​projected onto the magnetic field. , For the first Magnetic field projection observations of each magnetic field sensing unit; superscript It is the transpose symbol; The response matrix is ​​a multi-chip non-square matrix composed of the response coefficients of each magnetic field. ; The vector of the current to be inverted, , For the first The current to be inverted in the conductor under test; This is the measurement error vector; The weighted least squares algorithm with a regularization term is used to solve the problem, and the inverse objective function is: In the formula, This is the current estimation vector after inversion; This is the weight matrix. , Represents a diagonal matrix. For the first The weighting coefficients corresponding to each integrated measurement chip; For regularization parameters; This is the regularization matrix; It is an L2 norm.

4. The method for measuring conductor current based on a multi-sensor array according to claim 3, characterized in that, Weighting coefficient The calculation formula is: In the formula, For the first The function of weighting coefficients corresponding to each integrated measurement chip. For the first Noise uncertainty of a single magnetic field sensing unit; For the first Positioning uncertainty of each positioning unit; For the first The attitude uncertainty of each attitude sensing unit; For the first The structural calibration residual error of the integrated measurement chip.

5. The method for measuring conductor current based on a multi-sensor array according to claim 4, characterized in that, In step S7, after obtaining the estimated current value of the conductor under test, the following steps are also included: Based on the calculated current estimate and the measurement system equations, the inversion residuals corresponding to each integrated measurement chip are calculated using the following formula: In the formula, For the first The inversion residuals corresponding to each integrated measurement chip; For the first The estimated current value of the conductor to be tested; Determine whether each inversion residual exceeds the set residual threshold. When the inversion residual corresponding to a certain integrated measurement chip exceeds the threshold, assign a reduced weight coefficient to the chip, or remove it from the multi-chip overdetermined observation system and re-execute the weighted least squares algorithm for inversion calculation.

6. The method for measuring conductor current based on a multi-sensor array according to claim 3, characterized in that, When the current in the conductor under test is alternating current, the conductor current measurement method further includes: By repeatedly performing the inversion solution at multiple consecutive sampling times, the instantaneous current sequence of each conductor under test is obtained: In the formula, For each conductor to be tested on the 1st The instantaneous current sequence at each sampling time; For the first The test conductor is in the first... The estimated instantaneous current at each sampling time; Based on the instantaneous current sequence, the AC characteristic parameters of each conductor under test are calculated, and the AC characteristic parameters include at least one of effective value, peak value and phase.

7. The method for measuring conductor current based on a multi-sensor array according to claim 1, characterized in that, In step S2, the following evaluation function is used to select the optimal chip placement combination for actual observation from several integrated measurement chips: In the formula, The optimal chip placement combination; This refers to one type of chip arrangement combination; Chip placement and assembly The corresponding response matrix condition number; Indicates the arrangement of chips The inversion error is predicted by a variety of factors. Chip placement and assembly The corresponding deployment cost; These are the weighting coefficients.

8. The method for measuring conductor current based on a multi-sensor array according to claim 1, characterized in that, In step S1, the insulation calibration probe with the positioning tag sequentially contacts multiple positions on the surface of the conductor to be tested and records the corresponding spatial contact points; when there is a fixed offset between the contact tip of the insulation calibration probe and the positioning tag, the recorded contact points are corrected in combination with the fixed offset to calculate the accurate actual contact point position.

9. A conductor current measurement system based on a multi-sensor array, characterized in that, Used to implement the wire current measurement method based on a multi-sensor array as described in any one of claims 1 to 8; The conductor current measurement system includes: Multiple positioning reference nodes are deployed around the measurement area and used to establish the global coordinate system; An insulating calibration probe with a positioning label attached to its surface is used to contact the surface of the conductor under test to generate coordinate points for fitting. Insulation measuring bracket, including several mounting positions; The distributed measurement array consists of multiple integrated measurement chips, with the number of integrated measurement chips exceeding the number of conductors whose current is to be inverted. Each integrated measurement chip contains a magnetic field sensing unit, a positioning unit, and an attitude sensing unit. Multiple integrated measurement chips are distributed and installed in at least one mounting grid on an insulated measurement bracket, with positional differences in the lateral, height, and depth directions. An evaluation function based on the condition number of the response matrix, the predicted inversion error, and the deployment cost is used to select the optimal combination of chips to participate in the observation from multiple candidate mounting grids on the insulated measurement bracket.

Images