A composite current sensor based on a combination of a TMR array and an open-close Rogowski coil

Abstract

The application discloses a composite current sensor based on a combination of a TMR array and an open-close Rogowski coil, and comprises an open-close magnetic gathering ring made of a high-magnetic-permeability nanocrystalline soft magnetic material, the inner wall of the magnetic gathering ring is symmetrically provided with a multi-channel TMR sensor array to form a differential detection structure, and a semicircular ring-shaped PCB Rogowski coil is arranged outside the magnetic gathering ring and is used for sensing a current change rate; the TMR array detects a local magnetic field distortion caused by an opening air gap in real time, a three-layer error model is established, an extended Kalman filter or a least square algorithm is combined to online estimate complex correction coefficients, the amplitude and phase of the output of the Rogowski coil are compensated, the closed-loop correction of the air gap error is realized, the problem of the decrease of the measurement accuracy of the traditional open-close current sensor caused by the air gap is solved, the DC / low-frequency response advantage of the TMR sensor and the wide frequency band characteristic of the Rogowski coil are fused, the precision can reach 0.2 level in the 0-100 A measurement range, the application is suitable for electric energy metering and power protection systems, and the application has the advantages of high precision, easy installation and strong anti-interference capability.

Description

Technical Field

[0001] This invention belongs to the field of power measurement technology, specifically relating to a composite current sensor based on a combination of a TMR array and an open / closed Rogowski coil. Background Technology

[0002] Traditional iron-core current transformers suffer from increased measurement errors at high currents due to core magnetic saturation, and the bandwidth of iron-core current transformers is limited, making them unsuitable for measuring rapidly changing currents. Open-type Rogowski coils, which are coreless, enable wide-bandwidth measurements and remain unsaturated under high currents, offering flexible structure and easy installation. However, their output signal is a first-order integral, requiring a high-precision integrator and being susceptible to environmental factors. Furthermore, leakage flux at the opening of the Rogowski coil leads to ratio and angle errors between the actual induced electromotive force and the theoretical value. In the existing technology, the calibration and compensation methods for Rogowski coils are not perfect enough, especially the online compensation for errors caused by open leakage magnetic field. On the other hand, TMR sensors have extremely high magnetic field sensitivity and good linearity, and their temperature stability is better than that of Hall sensors. In high-precision magnetic field measurement, magnetic flux concentrators, i.e. magnetic rings, are often used to enhance the sensor output. However, traditional Hall sensors require large magnetic rings and complex settings, which increases the size and cost. To address the shortcomings of existing technologies, such as large measurement errors and complex calibration of open-close Rogowski coils, this invention provides a composite current sensing based on a combination of a TMR array and an open-close Rogowski coil. Summary of the Invention

[0003] To address the aforementioned issues, this invention provides a composite current sensor based on a combination of a TMR array and an open / closed Rogowski coil. This method can estimate and correct errors by sensing the open leakage magnetic field distribution in real time, thereby improving the measurement accuracy and reliability of the open / closed current transformer.

[0004] To solve the above-mentioned technical problems, the technical solution adopted by the present invention is as follows: A composite current sensor based on a combination of a TMR array and an open-closed Rogowski coil includes a ring magnetic circuit assembly and an open-closed Rogowski coil surrounding the magnetic circuit assembly. The Rogowski coil is coaxially nested with the ring magnetic circuit assembly and physically isolated. The inner wall of the ring magnetic circuit assembly is symmetrically arranged with a multi-channel magnetic field sensor array. Each pair of symmetrically arranged TMRs forms a full-bridge differential output channel to suppress temperature drift and bias error. The multi-channel magnetic field sensor array and the output of the Rogowski coil are connected to the data acquisition and processing unit. The data acquisition and processing unit connects to the host computer via a communication interface to display data, store data, configure parameters, and generate alarms for abnormalities.

[0005] Preferably, the multi-channel magnetic field sensor array includes eight magnetic field sensors, i.e., an eight-channel magnetic field sensor array. The eight magnetic field sensors are distributed along the inner wall of the magnetic ring at angles of 0°, 45°, 90°, 135°, 180°, 225°, 270°, and 315°. The core function is to detect and amplify the magnetic signal at the air gap. The toroidal magnetic circuit assembly uses a magnetic ring composed of a high-permeability soft magnetic material. The magnetic ring material is nanocrystalline, and its core function is to concentrate the leakage magnetic field of the measured current. The physical parameters of the magnetic ring meet any of the following requirements: initial relative permeability ; Saturation magnetic induction ; Within the operating frequency band, the specific magnetic loss is lower than a predetermined threshold; The magnetic ring is preferably machined into two halves and equipped with a mechanical positioning tongue to achieve high repeatability positioning accuracy when closed. The magnetic field sensor is selected from the tunnel magnetoresistive (TMR) sensor.

[0006] In actual operation, the magnetic ring and the Rogowski coil open and close synchronously to clamp the conductor under test. The TMR channel and the Rogowski coil output magnetic and voltage signals to the data acquisition and processing unit, respectively. The processed current signal is uploaded to the host computer, effectively compensating for air gap error and enabling the sensor to achieve an accuracy of 0.2 class in the 0~100A measurement range.

[0007] The Rogowski coil consists of two semi-circular toroidal PCB windings. The Rogowski coil's number of turns, inner and outer diameters, thickness, and other parameters must meet the following values: Number of turns N=600; inner diameter of coil approximately 64mm; outer diameter of coil approximately 84mm; coil thickness approximately 6mm; The output voltage signal of the Rogowski coil is then input into the data acquisition and processing unit.

[0008] In actual operation, the magnetic ring and the Rogowski coil open and close synchronously to clamp the conductor under test. The TMR channel and the Rogowski coil output magnetic and voltage signals to the data acquisition and processing unit, respectively. The processed current signal is uploaded to the host computer, effectively compensating for air gap error and enabling the sensor to achieve an accuracy of 0.2 class in the 0~100A measurement range.

[0009] The data acquisition and processing unit includes a fundamental frequency extraction module, a temperature compensation module, a real-time estimation module, a correction coefficient calculation module, and a human-machine interaction alarm unit, with FPGA and embedded processor as the hardware core.

[0010] The fundamental frequency extraction module includes a second-order complementary filter bank, which filters the low-frequency component output by the multi-channel magnetic field sensor array and the high-frequency component output by the Rogowski coil, and performs phase alignment at a preset crossover frequency to extract a full-band current signal containing the power frequency fundamental frequency. The temperature compensation module is used to acquire the real-time ambient temperature, calculate the temperature compensation factor based on the pre-stored magnetoresistive temperature characteristic model, and correct the output sensitivity of the multi-channel magnetic field sensor array in real time. The magnetoresistive temperature characteristic model is derived based on Bloch's law, and the formula is as follows: (1); In formula (1): For sensor reference sensitivity, The material's temperature drift coefficient. The temperature is collected in real time; The real-time estimation module is configured to calculate the phase difference statistics between the multi-channel magnetic field sensor array and the Rogowski coil output signal. And utilize the preset phase-air gap sensitivity mapping relationship Output the estimated value of the equivalent air gap. ; The correction coefficient calculation module determines the corresponding amplitude correction coefficient and phase correction coefficient by calling a preset polynomial fitting function based on the estimated value of the equivalent air gap. The human-machine interaction alarm unit issues an alarm based on the air gap determination logic.

[0011] The real-time estimation module is configured to calculate the phase difference statistics between the multi-channel magnetic field sensor array and the Rogowski coil output signal. And utilize the preset phase-air gap sensitivity mapping relationship Output the estimated value of the equivalent air gap. ; The correction coefficient calculation module determines the corresponding amplitude correction coefficient and phase correction coefficient by calling a preset polynomial fitting function based on the estimated value of the equivalent air gap. The real-time estimation module stores a three-layer error model, including a magnetic circuit model describing the relationship between air gap magnetoresistance and magnetic focusing efficiency, an exponentially decaying field distribution model describing the distortion of the air gap edge field, and a signal model describing the differential output of the TMR array, as shown in the following expressions: (2); In formula (2): The magnetic focusing efficiency is defined as the actual field when the air gap exists. The ratio of the field to the ideal closed field , It is an equivalent length constant. This is the equivalent air gap length. Relative permeability; For a reference field without a magnetic ring, and These are empirical parameters, depending on g, The geometry of the magnetic ring is determined by experimental fitting. This is the TMR bridge differential output voltage. This is the bias voltage. For the bridge arm resistance, As the reference sensitivity, This refers to the sensitivity temperature drift coefficient; The model establishes the sensor output voltage. With air gap length and primary current Nonlinear observation function between; The real-time estimation module also employs the extended Kalman filter algorithm; During operation, the extended Kalman filter algorithm uses the current observation value and the previous state estimate value to iteratively update the air gap estimate value in the state vector by calculating the Jacobian matrix of the observation function. The data acquisition and processing unit uses the real-time updated air gap estimate. The system queries the pre-stored mutual inductance-air gap characteristic curve and generates complex correction coefficients to dynamically compensate the output of the Rogowski coil, thereby eliminating the ratio difference and angle difference caused by the installation air gap.

[0012] The real-time estimation module also adjusts the correction coefficient during online operation. Perform interpolation and update in real time; The correction coefficient In complex form The magnitude function With phase function During the pre-configuration and calibration phase of the multi-channel magnetic field sensor array, the standard currents under multiple preset discrete air gap openings are measured on a bench and interpolated in a polynomial manner and stored in the non-volatile memory of the data acquisition and processing unit for use by the correction coefficient calculation module during runtime.

[0013] The data acquisition and processing unit compensates the output of the Rogowski coil using amplitude correction coefficients and phase correction coefficients, and then performs band fusion to generate the final measurement result. The output of the Rogowski coil is compensated using amplitude correction coefficients and phase correction coefficients, as shown in the following formula: (3); In formula (3): To calibrate the sensor output current, This is the induced voltage output by the Rogowski coil.

[0014] Band fusion is used to generate the final measurement results, using the following formula: (4); In equation (4): , These are the transfer functions of a second-order Butterworth low-pass filter and a high-pass filter, respectively. The relationship between angular frequency and cutoff frequency is as follows: (5); This configuration ensures performance across the entire frequency band. To achieve perfect energy complementarity; Final output frequency domain current for: (6); In formula (6): The TMR output current corrected for the temperature compensation module.

[0015] The output of the magnetic field sensor array is synthesized in the form of a differential combination of the left and right half-rings, which is used to directly trigger the air gap determination logic. The differential combination formula is as follows: (7); In equation (7): The output voltage difference of the TMR sensor array. The sum of the output voltage scalars of the left half-loop TMR array, The sum of the output voltage scalars of the right half-ring TMR array.

[0016] The air gap determination logic is as follows: The synthesized differential voltage ΔV is monitored in real time. When ΔV exceeds the preset first threshold, the human-machine interaction alarm unit sends a first instruction to the host computer to issue a prompt that the ring magnetic circuit assembly is not clamped. When ΔV exceeds the preset second threshold and the confidence level of the air gap estimation output by the real-time estimation module is greater than the preset confidence threshold, a second instruction is sent to the host computer to trigger a fault alarm. The first threshold and the second threshold are calculated based on the differential voltage-air gap characteristic curve obtained during the pre-configuration calibration stage and in combination with the preset safety margin coefficient. They are stored in the register of the data acquisition and processing unit for easy retrieval.

[0017] The main beneficial effects of this invention are as follows: 1) By optimizing the circuit compensation design, the system output accuracy reaches the 0.2S level, which is significantly better than the traditional open-type Rogowski coil current transformer. Specifically, compared with the uncompensated Rogowski coil, this invention can reduce the ratio error by several times and significantly reduce the phase error, especially in high-current pulse measurement scenarios.

[0018] 2) This invention can significantly reduce the ratio error and optimize the phase error, ensuring that the system maintains high accuracy over a wide dynamic range. This is due to the application of a compensation algorithm, which keeps the error within high standard limits.

[0019] 3) The system can accurately monitor currents from 0 to 100A and currents rich in low-frequency components, fully meeting the application requirements of power metering and relay protection, and improving the practicality and reliability of the system.

[0020] 4) The present invention adopts a busbar-type magnetic ring and open coil structure, which is easy to install and can be directly and seamlessly matched with existing high-voltage busbars and conductors, simplifying the engineering integration process.

[0021] 5) The magnetic ring is preferably made of nanocrystalline material, which can withstand the strong magnetic field generated by the large current and ensure that the sensor operates in the linear region; at the same time, the TMR array provides high-sensitivity magnetic field measurement and effectively suppresses external interference and temperature drift.

[0022] 6) By combining model calibration algorithms, the measurement error is ensured to remain within the specified limits under wide temperature, wide frequency band and high dynamic range conditions, thus improving the environmental adaptability of the system.

[0023] 7) This invention features a compact structure, flexible installation, and ease of integration and calibration. Its total weight does not exceed 2kg, demonstrating significant technological advancement and application prospects compared to traditional solutions. Attached Figure Description

[0024] The present invention will be further described below with reference to the accompanying drawings and embodiments.

[0025] Figure 1 This is a schematic diagram of the openable Rogowski-TMR composite current transformer structure of the present invention; Figure 2 The simulation results show the variation of the differential voltage ΔV of the left and right half-rings of the eight-channel TMR array with the air gap g. Figure 3 Comparison of the ratio difference curves of the Rogowski coil before and after compensation under different air gap g openings; Figure 4 Comparison of Rogowski coil phase error before and after compensation as a function of air gap g; Figure 5 A comparison chart showing the real-time estimation capability of the extended Kalman filter algorithm for the air gap g.

[0026] In the diagram: 1. Ring magnetic circuit assembly; 2. Rogowski coil; 3. Multi-channel magnetic field sensor array; 4. Data acquisition and processing unit; 5. Host computer. Detailed Implementation

[0027] Example 1: A composite current sensor based on a combination of a TMR array and an open / closed Rogowski coil, comprising a ring magnetic circuit assembly 1 and an open / closed Rogowski coil 2 surrounding the magnetic circuit assembly; The inner wall of the annular magnetic circuit assembly is symmetrically arranged with a multi-channel magnetic field sensor array 3. Each pair of symmetrically arranged TMRs forms a full-bridge differential output channel to suppress temperature drift and bias error. The output of the multi-channel magnetic field sensor array 3 and the Rogowski coil 2 are connected to the data acquisition and processing unit 4.

[0028] The multi-channel magnetic field sensor array 3 includes multiple magnetic field sensors; The annular magnetic circuit assembly 1 consists of two semi-arc magnetic circuit assemblies and has an openable and closable structure to clamp the conductor under test.

[0029] Preferably, the multi-channel magnetic field sensor array 3 includes eight magnetic field sensors, that is, an eight-channel magnetic field sensor array; The toroidal magnetic circuit assembly uses a magnetic ring composed of VITROPERM500F nanocrystalline soft magnetic material, which has excellent performance and a high saturation magnetic flux density. Initial permeability Furthermore, it has very low eddy current loss, which enables the magnetic ring to have a significant accumulation effect on the leakage flux generated by the flowing current. The outer diameter of this series of magnetic rings ranges from 10mm to 60mm, and they can be placed inside Rogowski coils with an inner diameter of 64mm. They can also be made into a two-part structure that can be opened and closed, making it easy to clamp them under power.

[0030] The magnetic ring is preferably machined into two halves and equipped with a mechanical positioning tongue to achieve high repeatability positioning accuracy when closed. The magnetic field sensor is selected from the Tunnel Magnetoresistive (TMR) sensor, specifically the TMR2152S TMR sensor chip with a high dynamic range. It features a wide dynamic range of ±500Gs, excellent temperature stability, and 1% nonlinearity. Furthermore, the SOT23-5 package facilitates integration onto a PCB-type Rogowski coil. Two TMRs arranged symmetrically form a full-bridge differential output channel to suppress temperature drift and bias errors.

[0031] The Rogowski coil 2 consists of two semi-circular ring-shaped PCB windings. The Rogowski coil's number of turns, inner and outer diameters, thickness, and other parameters must meet the following values: Number of turns N=600; inner diameter of coil approximately 64mm; outer diameter of coil approximately 84mm; coil thickness approximately 6mm; The voltage signal output by the Rogowski coil 2 is then input into the data acquisition and processing unit 4.

[0032] The data acquisition and processing unit 4 estimates the equivalent air gap at the opening and closing of the annular magnetic circuit based on the output of the multi-channel magnetic field sensor array 3, and performs online correction of the amplitude and phase of the output of the Rogowski coil 2 according to the estimation result, thereby obtaining the final measured current value output. The data acquisition and processing unit 4 includes a fundamental frequency extraction module, a temperature compensation module, a real-time estimation module, a correction coefficient calculation module, and a human-machine interaction alarm unit.

[0033] The fundamental frequency extraction module includes a second-order complementary filter bank, which filters the low-frequency component output by the multi-channel magnetic field sensor array 3 and the high-frequency component output by the Rogowski coil 2, and performs phase alignment at a preset crossover frequency to extract a full-band current signal containing the power frequency fundamental frequency. The temperature compensation module is used to acquire the real-time ambient temperature, calculate the temperature compensation factor based on the pre-stored magnetoresistive temperature characteristic model, and correct the output sensitivity of the multi-channel magnetic field sensor array 3 in real time. The magnetoresistive temperature characteristic model is derived based on Bloch's law, and the formula is as follows: (1); In formula (1): For sensor reference sensitivity, The material's temperature drift coefficient. The temperature is collected in real time; The real-time estimation module is configured to calculate the phase difference statistics between the multi-channel magnetic field sensor array 3 and the output signal of the Rogowski coil 2. And utilize the preset phase-air gap sensitivity mapping relationship Output the estimated value of the equivalent air gap. ; The correction coefficient calculation module determines the corresponding amplitude correction coefficient and phase correction coefficient by calling a preset polynomial fitting function based on the estimated value of the equivalent air gap. The human-machine interaction alarm unit issues an alarm based on the air gap determination logic.

[0034] The real-time estimation module stores a three-layer error model, including a magnetic circuit model describing the relationship between air gap magnetoresistance and magnetic focusing efficiency, an exponentially decaying field distribution model describing the distortion of the air gap edge field, and a signal model describing the differential output of the TMR array, as shown in the following expressions: (2); In formula (2): The magnetic focusing efficiency is defined as the actual field when the air gap exists. The ratio of the field to the ideal closed field , It is an equivalent length constant. This is the equivalent air gap length. Relative permeability; For a reference field without a magnetic ring, and These are empirical parameters, depending on g, The geometry of the magnetic ring is determined by experimental fitting. This is the TMR bridge differential output voltage. This is the bias voltage. For the bridge arm resistance, As the reference sensitivity, This refers to the sensitivity temperature drift coefficient; The model establishes the sensor output voltage. With air gap length and primary current Nonlinear observation function between; The real-time estimation module also employs the extended Kalman filter algorithm; During operation, the extended Kalman filter algorithm uses the current observation value and the previous state estimate value to iteratively update the air gap estimate value in the state vector by calculating the Jacobian matrix of the observation function. The data acquisition and processing unit 4 uses the real-time updated air gap estimate. The system queries the pre-stored mutual inductance-air gap characteristic curve and generates complex correction coefficients to dynamically compensate the output of the Rogowski coil, thereby eliminating the ratio difference and angle difference caused by the installation air gap.

[0035] The real-time estimation module also adjusts the correction coefficient during online operation. Perform interpolation and update in real time; The correction coefficient In complex form The magnitude function With phase function During the pre-configuration calibration stage of the multi-channel magnetic field sensor array 3, the standard current under multiple preset discrete air gap openings is obtained by bench measurement and interpolated in a polynomial manner and stored in the non-volatile memory of the data acquisition and processing unit for the correction coefficient calculation module to call during operation.

[0036] After the data acquisition and processing unit 4 compensates the output of the Rogowski coil 2 using the amplitude correction coefficient and the phase correction coefficient, it performs frequency band fusion to generate the final measurement result. The output of the Rogowski coil 2 is compensated using amplitude correction coefficients and phase correction coefficients, as shown in the following formula: (3); In formula (3): To calibrate the sensor output current, This is the induced voltage output by the Rogowski coil.

[0037] Band fusion is used to generate the final measurement results, using the following formula: (4); In equation (4): , These are the transfer functions of a second-order Butterworth low-pass filter and a high-pass filter, respectively. The relationship between angular frequency and cutoff frequency is as follows: (5); This configuration ensures performance across the entire frequency band. To achieve perfect energy complementarity; Final output frequency domain current for: (6); In formula (6): The TMR output current corrected for the temperature compensation module.

[0038] The output of the multi-channel magnetic field sensor array 3 is synthesized in the form of a differential combination of the left and right half-rings and is used to directly trigger the air gap determination logic. The differential combination formula is as follows: (7); In equation (7): The output voltage difference of the TMR sensor array. The sum of the output voltage scalars of the left half-loop TMR array, The sum of the output voltage scalars of the right half-ring TMR array.

[0039] The air gap determination logic is as follows: Real-time monitoring of the synthesized differential voltage Δ V When Δ V When the preset first threshold is exceeded, the human-machine interaction alarm unit sends a first instruction to the host computer to issue a prompt that the ring magnetic circuit assembly is not clamped. When Δ V When the air gap estimation confidence level output by the real-time estimation module exceeds the preset second threshold and is greater than the preset confidence threshold, the human-machine interaction alarm unit sends a second instruction to the host computer to trigger a fault alarm. The first threshold and the second threshold are calculated based on the differential voltage-air gap characteristic curve obtained during the pre-configuration calibration stage and in combination with the preset safety margin coefficient. They are stored in the register of the data acquisition and processing unit for easy retrieval.

[0040] The specific operating process of this invention is as follows: First, the model parameters of the composite sensor are obtained through offline calibration; After the device is mounted, the data acquisition and processing section receives, filters / integrates, compensates, and corrects the temperature of the TMR and Rogowski signals. Finally, the processed current value, which is closer to the real value, is output to the host computer through frequency band allocation. The host computer displays the current data of the busbar segment. If the air gap is too large during clamping and exceeds the first threshold, an alarm signal will be sent directly to the host computer via the FPGA in the TMR signal receiving stage. If the air gap exceeds the second threshold, a strong alarm signal will be sent to prompt the staff to re-clamp the device.

Claims

1. A composite current sensor based on a combination of a TMR array and an open / closed Rogowski coil, characterized in that: It includes a toroidal magnetic circuit assembly (1) and an open-closed Rogowski coil (2) surrounding the magnetic circuit assembly; The inner wall of the annular magnetic circuit assembly is symmetrically arranged with a multi-channel magnetic field sensor array (3), and the output of the multi-channel magnetic field sensor array (3) and the Rogowski coil (2) are connected to the data acquisition and processing unit (4). The data acquisition and processing unit (4) is connected to the host computer (5) through a communication interface to realize data display, storage, parameter configuration and abnormal alarm.

2. The composite current sensor based on a combination of a TMR array and an open / closed Rogowski coil according to claim 1, characterized in that: The multi-channel magnetic field sensor array (3) includes multiple magnetic field sensors; The annular magnetic circuit assembly (1) consists of two semi-arc magnetic circuit assemblies and is an openable structure for clamping the conductor under test.

3. A composite current sensor based on a combination of a TMR array and an open / closed Rogowski coil as described in claim 1, characterized in that: The Rogowski coil (2) consists of two semi-circular toroidal PCB windings; The output voltage signal of the Rogowski coil (2) is then input into the data acquisition and processing unit (4).

4. A composite current sensor based on a combination of a TMR array and an open / closed Rogowski coil as described in claim 1, characterized in that: The data acquisition and processing unit (4) estimates the equivalent air gap at the opening and closing of the annular magnetic circuit based on the output of the multi-channel magnetic field sensor array (3), and performs online correction of the amplitude and phase of the output of the Rogowski coil (2) according to the estimation result, so as to obtain the final measured current value output. The data acquisition and processing unit (4) includes a fundamental wave extraction module, a temperature compensation module, a real-time estimation module, a correction coefficient calculation module, a human-machine interaction alarm unit, and uses FPGA and embedded processor as the hardware core.

5. A composite current sensor based on a combination of a TMR array and an open / closed Rogowski coil as described in claim 4, characterized in that: The fundamental wave extraction module includes a second-order complementary filter bank. The second-order complementary filter bank filters the low-frequency component output by the multi-channel magnetic field sensor array (3) and the high-frequency component output by the Rogowski coil (2), and performs phase alignment at a preset crossover frequency to extract a full-band current signal containing the power frequency fundamental wave. The temperature compensation module is used to obtain the real-time ambient temperature, calculate the temperature compensation factor based on the pre-stored magnetoresistive temperature characteristic model, and correct the output sensitivity of the multi-channel magnetic field sensor array (3) in real time. The magnetoresistive temperature characteristic model is derived based on Bloch's law, and the formula is as follows: （1）； In formula (1): For sensor reference sensitivity, The material's temperature drift coefficient. The temperature is collected in real time; The real-time estimation module is configured to calculate the phase difference statistics between the multi-channel magnetic field sensor array (3) and the output signal of the Rogowski coil (2). And utilize the preset phase-air gap sensitivity mapping relationship Output the estimated value of the equivalent air gap. ; The correction coefficient calculation module determines the corresponding amplitude correction coefficient and phase correction coefficient by calling a preset polynomial fitting function based on the estimated value of the equivalent air gap. The human-machine interaction alarm unit issues an alarm based on the air gap determination logic.

6. A composite current sensor based on a combination of a TMR array and an open / closed Rogowski coil as described in claim 5, characterized in that: The real-time estimation module stores a three-layer error model, including a magnetic circuit model describing the relationship between air gap magnetoresistance and magnetic focusing efficiency, an exponentially decaying field distribution model describing the distortion of the air gap edge field, and a signal model describing the differential output of the TMR array, as shown in the following expressions: （2）； In formula (2): The magnetic focusing efficiency is defined as the actual field when the air gap exists. The ratio of the field to the ideal closed field , It is an equivalent length constant. This is the equivalent air gap length. Relative permeability; For a reference field without a magnetic ring, and These are empirical parameters, depending on g, The geometry of the magnetic ring is determined by experimental fitting. This is the TMR bridge differential output voltage. This is the bias voltage. For the bridge arm resistance, As the reference sensitivity, This refers to the sensitivity temperature drift coefficient; The model establishes the sensor output voltage. With air gap length and primary current Nonlinear observation function between; The real-time estimation module also employs the extended Kalman filter algorithm; During operation, the extended Kalman filter algorithm uses the current observation value and the previous state estimate value to iteratively update the air gap estimate value in the state vector by calculating the Jacobian matrix of the observation function. The data acquisition and processing unit (4) uses the real-time updated air gap estimate. The system queries the pre-stored mutual inductance-air gap characteristic curve and generates complex correction coefficients to dynamically compensate the output of the Rogowski coil, thereby eliminating the ratio difference and angle difference caused by the installation air gap.

7. A composite current sensor based on a combination of a TMR array and an open / closed Rogowski coil as described in claim 6, characterized in that: The real-time estimation module also adjusts the correction coefficient during online operation. Perform interpolation and update in real time; The correction coefficient In complex form The magnitude function With phase function In the pre-configuration calibration stage of the multi-channel magnetic field sensor array (3), the standard current under multiple preset discrete air gap openings is obtained by bench measurement and interpolated in a polynomial manner and stored in the non-volatile memory of the data acquisition and processing unit for the correction coefficient calculation module to call during operation.

8. A composite current sensor based on a combination of a TMR array and an open / closed Rogowski coil according to claim 7, characterized in that: The data acquisition and processing unit (4) compensates the output of the Rogowski coil (2) using the amplitude correction coefficient and the phase correction coefficient, and then performs band fusion to generate the final measurement result; The output of the Rogowski coil (2) is compensated using amplitude correction coefficients and phase correction coefficients, and the specific formulas are as follows: （3）； In formula (3): To calibrate the sensor output current, This is the induced voltage output by the Rogowski coil. Band fusion is used to generate the final measurement results, using the following formula: （4）； In equation (4): , These are the transfer functions of a second-order Butterworth low-pass filter and a high-pass filter, respectively. The relationship between angular frequency and cutoff frequency is as follows: （5）； This configuration ensures coverage across the entire frequency band. To achieve perfect energy complementarity; Final output frequency domain current for: （6）； In formula (6): The TMR output current corrected for the temperature compensation module.

9. A composite current sensor based on a combination of a TMR array and an open / closed Rogowski coil according to claim 1, characterized in that: The output of the multi-channel magnetic field sensor array (3) is synthesized in the form of a differential between the left and right half-rings and is used to directly trigger the air gap determination logic. The differential synthesis formula is as follows: （7）； In equation (7): The output voltage difference of the TMR sensor array. The sum of the output voltage scalars of the left half-loop TMR array, The sum of the output voltage scalars of the right half-ring TMR array.

10. A composite current sensor based on a combination of a TMR array and an open / closed Rogowski coil according to claim 9, characterized in that: The air gap determination logic is as follows: The synthesized differential voltage ΔV is monitored in real time. When ΔV exceeds the preset first threshold, the human-machine interaction alarm unit sends a first instruction to the host computer (5) to issue a prompt that the ring magnetic circuit assembly is not clamped. When ΔV exceeds the preset second threshold and the confidence level of the air gap estimation output by the real-time estimation module is greater than the preset confidence threshold, the human-machine interaction alarm unit sends a second instruction to the host computer (5) to trigger a fault alarm. The first threshold and the second threshold are calculated based on the differential voltage-air gap characteristic curve obtained during the pre-configuration calibration stage and in combination with the preset safety margin coefficient. They are stored in the register of the data acquisition and processing unit for easy retrieval.

Images