Unmanned helicopter magnetic interference real-time compensation method and system based on improved Kalman filtering

Abstract

The invention provides an unmanned helicopter magnetic interference real-time compensation method and system based on improved Kalman filtering, and relates to the technical field of unmanned helicopter navigation and control. The response processing comprises flight control program initialization, attitude parameter loading and sensor self-inspection operation. Time synchronization processing is carried out on magnetic field combined sensing data collected in the flight process of the unmanned helicopter to form a multi-source flight data set, and a unified timestamp calibration mechanism is adopted in the time synchronization processing. According to the unmanned helicopter magnetic interference real-time compensation method and system based on improved Kalman filtering, the improved Kalman filtering algorithm is adopted to compensate the magnetic interference of the unmanned helicopter in real time. Through state estimation and dynamic weight adjustment, a kinetic model and a magnetic field distribution model of the unmanned helicopter are combined, magnetic field error correction in the flight process is achieved, and accurate control over the flight attitude is guaranteed.

Description

Technical Field

[0001] This invention relates to the field of unmanned aerial vehicle (UAV) navigation and control technology, specifically to a method and system for real-time magnetic interference compensation for unmanned helicopters based on improved Kalman filtering. Background Technology

[0002] Existing unmanned helicopters performing aeromagnetic surveys are often affected by magnetic interference generated by their airframe structure, motors, power systems, and electronic equipment. This magnetic interference superimposed on the geomagnetic signal, distorting measurement data and affecting the accuracy of geomagnetic anomaly analysis and interpretation. Traditional aeromagnetic compensation techniques mostly rely on ground calibration, requiring operators to manually control the helicopter to perform preset attitude maneuvers and collect magnetometer data using ground equipment. Compensation parameters are typically calculated offline and loaded into the flight control system before the mission, making the process cumbersome and dependent on human experience. Some improved solutions attempt to embed real-time compensation algorithms into the flight control system, but most use fixed parameters or linear models, failing to fully consider the dynamic changes in flight attitude and the nonlinear characteristics of the environmental magnetic field, making it difficult to maintain compensation stability in complex flight environments.

[0003] During dynamic flight, traditional Kalman filtering algorithms rely on fixed noise model settings and lack adaptive adjustment mechanisms. When attitude angles, velocity, or acceleration change significantly, filter estimation errors increase, leading to delayed or even distorted compensation results. This results in drifting and jumps in magnetic measurement data, affecting the continuous analysis of geomagnetic anomalies. Particularly when performing missions in high-altitude, coastal, or strong magnetic gradient areas, the external magnetic field fluctuates dramatically, and static compensation parameters cannot adapt to complex environmental changes. Existing systems generally suffer from slow response, computational delays, and parameter mismatches, making real-time identification and dynamic compensation of magnetic interference during flight impossible. This problem directly limits the application effectiveness of unmanned helicopters in high-precision aeromagnetic surveying and scientific exploration. Summary of the Invention

[0004] To address the shortcomings of existing technologies, this invention provides a real-time magnetic interference compensation method and system for unmanned helicopters based on an improved Kalman filter. The technical problem this invention aims to solve is: how to address the attitude instability and insufficient control precision of unmanned helicopters caused by magnetic interference during flight by combining an improved Kalman filter algorithm with multi-source data synchronization and flight state adaptive compensation methods.

[0005] To achieve the above objectives, the present invention provides the following technical solution: a real-time magnetic interference compensation method for unmanned helicopters based on improved Kalman filtering, comprising: S1. The unmanned helicopter flight control system receives a trigger command and processes it to form a compensation start signal. The response processing includes flight control program initialization, attitude parameter loading, and sensor self-test operation. S2. Time synchronization processing is performed on the magnetic field joint sensing data collected during the flight of the unmanned helicopter to form a multi-source flight dataset. The time synchronization processing adopts a unified timestamp calibration mechanism. S3. Perform state estimation processing on the multi-source flight dataset to form magnetic interference estimation results. The state estimation processing adopts an improved Kalman filter algorithm. The improved Kalman filter algorithm adopts a joint state space equation, which includes an unmanned helicopter dynamics model and a magnetic field distribution model. S4. Perform adaptive compensation processing on the magnetic interference estimation results to form magnetic field correction data. The adaptive compensation processing includes adjusting the compensation coefficient and calculating the compensation parameters. S5. Perform difference evaluation processing on the magnetic field correction data to form a compensation feedback update result. The difference evaluation processing includes magnetic field strength fluctuation comparison and gradient feature analysis. Transmit the compensation feedback update result to the flight control system. The compensation feedback update result updates the parameters of the improved Kalman filter algorithm to form a closed-loop compensation mechanism.

[0006] Preferably, the flight control program initialization includes loading flight control parameters, setting initial values ​​for the attitude control loop PID, and automatic correction of magnetometer bias. The attitude parameter loading includes attitude angle calculation, and the sensor self-test operation includes amplitude and noise characteristic detection.

[0007] Preferably, the combined magnetic field sensing data includes magnetometer data, accelerometer data, gyroscope data, and GPS data.

[0008] Preferably, the improved Kalman filter algorithm includes prediction, updating, and adaptive adjustment. The joint state-space equation adopts a dynamic weighting and nonlinear correction coupling formula, the expression of which is: .

[0009] in, The result is a dimensionless state estimate for the next time step. The current state estimation vector, dimensionless. For observing residuals, dimensionless, This is the dynamic gain matrix, dimensionless, with values ​​ranging from 0 to 1. The weight matrix is ​​a linear, dimensionless matrix. The weight matrix is ​​non-linear and dimensionless. Let L be the square of the L2 norm of the residual vector, which is dimensionless. It is a hyperbolic tangent function, dimensionless, with an amplitude range of [-1, 1].

[0010] Preferably, the unmanned helicopter dynamics model outputs the current state estimation vector, which includes attitude angle state variables and angular velocity state variables. The magnetic field distribution model outputs the observation residual, which includes a geomagnetic reference vector, a direction cosine matrix, and magnetic disturbance components. The unmanned helicopter dynamics model and the magnetic field distribution model are bidirectionally coupled.

[0011] Preferably, the compensation coefficient is dynamically corrected based on the attitude angular velocity, acceleration, and magnetic disturbance rate of the unmanned helicopter, and the compensation parameter is the geomagnetic direction vector of the unmanned helicopter dynamic model and the magnetic field distribution model.

[0012] Preferably, the difference assessment process employs a nonlinear feedback adjustment formula, the calculation formula of which is: .

[0013] in, The feedback gain adjustment is dimensionless. The gain scaling factor is dimensionless. The sensitivity coefficient is dimensionless. The magnetic field gradient vector at the current moment is dimensionless. The magnetic field gradient vector from the previous moment is dimensionless. It is a nonlinear suppression function, dimensionless.

[0014] Preferably, the compensation feedback update result includes magnetic field strength fluctuation data, spatial gradient feature information, parameter update amount, and closed-loop control signal.

[0015] Preferably, the compensation feedback update result is transmitted using a UART interface and an SPI bus. The UART interface transmits the parameter update amount and the closed-loop control signal, and the SPI bus transmits the magnetic field strength fluctuation data and the spatial gradient characteristic information.

[0016] A real-time magnetic interference compensation system for unmanned helicopters based on improved Kalman filtering includes: The sensor module collects magnetic field and attitude information during the flight of the unmanned helicopter and outputs joint sensing data of flight attitude and magnetic field. The sensor module includes a three-axis magnetometer, a three-axis accelerometer, a three-axis gyroscope and a GPS module. The flight control main control unit performs time synchronization and state estimation processing on the joint sensing data of flight attitude and magnetic field to form a magnetic interference estimation result. The time synchronization and state estimation processing adopts an improved Kalman filter algorithm. The improved Kalman filter algorithm establishes a joint state space equation based on the dynamic model of the unmanned helicopter and the magnetic field distribution model. The storage unit stores the flight control program, aeromagnetic compensation algorithm, historical compensation parameters, and flight data; A communication unit that transmits the magnetic interference estimation result, the communication unit including a wireless communication module and a data interface; An execution unit is provided, which adjusts the flight attitude of the unmanned helicopter based on the magnetic interference estimation result. The execution unit includes a motor controller and a servo controller.

[0017] This invention provides a method and system for real-time magnetic interference compensation for unmanned helicopters based on improved Kalman filtering. It has the following beneficial effects: This invention relates to a real-time magnetic interference compensation method and system for unmanned helicopters based on an improved Kalman filter. By employing an improved Kalman filter algorithm, the method compensates for magnetic interference in real time. Through state estimation and dynamic weight adjustment, it combines the dynamic model and magnetic field distribution model of the unmanned helicopter to achieve magnetic field error correction during flight, ensuring precise control of flight attitude.

[0018] The improved Kalman filter algorithm and adaptive compensation processing mechanism adopted dynamically correct the compensation coefficients and parameters, and adjust the magnetic field correction data in real time, thereby improving the adaptability of flight status and the stability of the system, enhancing the accuracy and reliability of magnetic interference compensation, and ensuring the efficient flight of the unmanned helicopter in complex environments. Attached Figure Description

[0019] Figure 1 This is a flowchart illustrating the overall process of the method of the present invention. Figure 2 This is a schematic diagram of the system structure of the present invention; Figure 3 This is a flowchart of the state estimation process of the present invention; Figure 4 This is the adaptive compensation processing flow of the present invention; Figure 5 This is a flowchart of the difference assessment and feedback update process for this invention. Detailed Implementation

[0020] The technical solutions in the embodiments of the present invention have been clearly and completely described. 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.

[0021] Example 1 like Figure 1-5As shown, this embodiment of the invention provides a real-time magnetic interference compensation method for unmanned helicopters based on an improved Kalman filter, including: S1, processing the trigger command received by the unmanned helicopter flight control system to form a compensation start signal; the response processing includes flight control program initialization, attitude parameter loading, and sensor self-testing. Flight control program initialization includes loading flight control parameters, setting initial values ​​for the attitude control loop PID, and automatic correction of magnetometer bias; attitude parameter loading includes attitude angle calculation; and sensor self-testing includes amplitude and noise characteristic detection.

[0022] S2. Time synchronization processing is performed on the magnetic field joint sensing data collected during the flight of the unmanned helicopter to form a multi-source flight dataset. The time synchronization processing adopts a unified timestamp calibration mechanism. The magnetic field joint sensing data includes magnetometer data, accelerometer data, gyroscope data, and GPS data.

[0023] S3. State estimation processing is performed on the multi-source flight dataset to generate magnetic interference estimation results. The state estimation process employs an improved Kalman filter algorithm, which uses a joint state-space equation. This joint state-space equation includes the unmanned helicopter dynamics model and the magnetic field distribution model. The improved Kalman filter algorithm includes prediction, updating, and adaptive adjustment. The joint state-space equation uses a dynamic weighting and nonlinear correction coupling formula, the expression of which is: .

[0024] in, The result is a dimensionless state estimate for the next time step. The current state estimation vector, dimensionless. For observing residuals, dimensionless, This is the dynamic gain matrix, dimensionless, with values ​​ranging from 0 to 1. The weight matrix is ​​a linear, dimensionless matrix. The weight matrix is ​​non-linear and dimensionless. Let L be the square of the L2 norm of the residual vector, which is dimensionless. It is a hyperbolic tangent function, dimensionless, with an amplitude range of [-1, 1].

[0025] The unmanned helicopter dynamics model outputs the current state estimation vector, which includes attitude angle state variables and angular velocity state variables. The magnetic field distribution model outputs the observation residuals, which include the geomagnetic reference vector, direction cosine matrix and magnetic disturbance components. The unmanned helicopter dynamics model and the magnetic field distribution model are bidirectionally coupled.

[0026] S4. Adaptive compensation processing based on flight state is performed on the magnetic disturbance estimation results to generate magnetic field correction data. The adaptive compensation processing includes adjusting the compensation coefficients and calculating the compensation parameters. The compensation coefficients are dynamically corrected based on the unmanned helicopter's attitude angular velocity, acceleration, and rate of change of magnetic disturbance. The compensation parameters are the geomagnetic direction vectors of the unmanned helicopter's dynamic model and magnetic field distribution model.

[0027] S5. The magnetic field correction data undergoes a difference assessment process to generate a compensation feedback update result. This process includes magnetic field strength fluctuation comparison and gradient feature analysis. The compensation feedback update result is then transmitted to the flight control system, where it updates the parameters of the improved Kalman filter algorithm, forming a closed-loop compensation mechanism. The difference assessment process employs a nonlinear feedback adjustment formula. The calculation formula for the nonlinear feedback adjustment formula is as follows: .

[0028] in, The feedback gain adjustment is dimensionless. The gain scaling factor is dimensionless. The sensitivity coefficient is dimensionless. The magnetic field gradient vector at the current moment is dimensionless. The magnetic field gradient vector from the previous moment is dimensionless. It is a nonlinear suppression function, dimensionless.

[0029] The compensation feedback update results include magnetic field strength fluctuation data, spatial gradient characteristic information, parameter update amounts, and closed-loop control signals. The compensation feedback update results are transmitted using a UART interface and an SPI bus. The UART interface transmits parameter update amounts and closed-loop control signals, while the SPI bus transmits magnetic field strength fluctuation data and spatial gradient characteristic information.

[0030] A real-time magnetic interference compensation system for unmanned helicopters based on improved Kalman filtering includes: The sensor module collects magnetic field and attitude information during the flight of the unmanned helicopter and outputs combined flight attitude and magnetic field sensing data. The sensor module includes a three-axis magnetometer, a three-axis accelerometer, a three-axis gyroscope, and a GPS module.

[0031] The flight control main control unit performs time synchronization and state estimation processing on the joint sensing data of flight attitude and magnetic field to form magnetic interference estimation results. The time synchronization and state estimation processing adopts an improved Kalman filter algorithm, which establishes a joint state space equation based on the dynamic model of the unmanned helicopter and the magnetic field distribution model.

[0032] The storage unit stores the flight control program, aeromagnetic compensation algorithm, historical compensation parameters, and flight data.

[0033] The communication unit transmits the magnetic interference estimation results and includes a wireless communication module and a data interface.

[0034] The execution unit adjusts the flight attitude of the unmanned helicopter based on the magnetic interference estimation results. The execution unit includes a motor controller and a servo controller.

[0035] This real-time magnetic interference compensation method for unmanned helicopters based on an improved Kalman filter achieves real-time compensation for magnetic interference. It features key characteristics such as dynamic adjustment of compensation coefficients, integration of multi-source data, and closed-loop feedback control. By dynamically correcting flight attitude, acceleration, and the rate of change of magnetic disturbance, the compensation process can adapt to changes in flight state, improving system stability and adaptability. The real-time magnetic interference compensation system for unmanned helicopters based on the improved Kalman filter continuously optimizes the compensation effect through real-time evaluation of magnetic field strength fluctuations and gradient characteristics, and utilizes multi-source sensor data to improve the accuracy of magnetic interference estimation, ensuring smooth and accurate flight.

[0036] Example 2 This embodiment presents a real-time magnetic interference compensation method and system for unmanned helicopters based on an improved Kalman filter. By processing the sensor data of the unmanned helicopter dimensionlessly, real-time compensation is achieved in magnetic interference environments to improve the flight stability of the aircraft. The specific implementation method is as follows: 1. Data Input Preparation During a particular flight, the sensor module of the unmanned helicopter collected the following data: Magnetometer data, in Gauss: 45.2, 37.8, 23.5. Accelerometer data, in m / s². 2 : 0.12, 0.05, -0.03. Gyroscope data, in rad / s: 0.05, -0.02, 0.08. GPS data, in degrees: 37.7749, -122.4194, 150.

[0037] 2. Dimensionless process Dimensionless data processing normalizes raw data, making it independent of physical units and adaptable to different scenarios. The maximum value normalization method is used, and the specific steps are as follows: Magnetometer data: The maximum geomagnetic field strength is 50 Gauss.

[0038]

[0039] Accelerometer data: Maximum gravitational acceleration is 9.8 m / s². 2 .

[0040] .

[0041] Gyroscope data: Maximum angular velocity is 10 rad / s.

[0042]

[0043] GPS data: Longitude range [-180°, 180°], latitude range [-90°, 90°], altitude range [0, 5000] meters.

[0044]

[0045]

[0046]

[0047] The normalized GPS data is as follows:

[0048] 3. Current state estimate The flight control system outputs the current state estimate, expressed as a dimensionless quantity:

[0049] 4. Calculation of observation residuals Based on the flight control system model, the geomagnetic reference vector at the current moment is obtained:

[0050] Calculate the observation residuals:

[0051]

[0052] 5. Dynamic gain matrix and weight matrix The flight control system uses the following values ​​to dynamically adjust the gain matrix and weight matrix: Dynamic gain matrix: The dynamic gain matrix is ​​used to balance the weights of prior estimates and current observations. It is adaptively adjusted based on the system's state estimation error using a Kalman filter algorithm. The current estimate has a high degree of confidence in previous estimates and is set based on empirical values. It is suitable for environments with low noise levels.

[0053] The linear weighting matrix is ​​a parameter used to balance the influence of the linear component in the Kalman filter on the state estimation. It is adjusted based on the magnitude of the observation residuals and the system's confidence in each sensor. The nonlinear weighting matrix controls the influence of the nonlinear adjustment component on the estimation results and is related to the system's nonlinear characteristics, the nonlinear factors of the observation errors, and the system's dynamic changes. (Setting...) , .

[0054] 6. Calculate the squared L2 norm of the residual vector. Calculate the squared L2 norm of the residual vector:

[0055] 7. Calculation of the hyperbolic tangent function Calculate the nonlinear adjustment term:

[0056] First, calculate :

[0057]

[0058]

[0059] Applying the hyperbolic tangent function The amplitude range is [-1, 1]:

[0060] 8. State estimation update The improved Kalman filter algorithm includes prediction, updating, and adaptive adjustment. The joint state-space equation adopts a dynamic weighting and nonlinear correction coupling formula, the expression of which is: .

[0061] in, The result is a dimensionless state estimate for the next time step. The current state estimation vector, dimensionless. For observing residuals, dimensionless, This is the dynamic gain matrix, dimensionless, with values ​​ranging from 0 to 1. The weight matrix is ​​a linear, dimensionless matrix. The weight matrix is ​​non-linear and dimensionless. Let L be the square of the L2 norm of the residual vector, which is dimensionless. It is a hyperbolic tangent function, dimensionless, with an amplitude range of [-1, 1].

[0062] The current state estimate is: .

[0063] Calculate state estimate update:

[0064]

[0065] 9. Output magnetic interference estimation results The magnetic interference estimation results are as follows:

[0066] The geomagnetic reference vector is estimated as follows: .

[0067] 10. Subsequent data transmission and compensation Magnetic interference estimation results: The magnetic interference estimation results will be transmitted to the flight control system via the wireless communication module.

[0068] Compensation parameters: The flight control system uses the estimated results to adjust the attitude and update the compensation coefficients.

[0069] The data will be continuously transmitted and updated in real time to ensure that the unmanned helicopter can fly stably in environments with strong magnetic interference.

[0070] Through the above steps, real-time estimation and compensation of magnetic interference are achieved. Normalized magnetometer, accelerometer, gyroscope, and GPS data ensure the applicability and reliability of the algorithm, enabling the system to operate stably under various flight conditions. The flight control system utilizes the real-time magnetic interference estimation results to precisely adjust the aircraft's attitude, ensuring flight stability and precise control even in environments with strong magnetic interference.

[0071] Example 3 This embodiment presents a real-time magnetic interference compensation method and system for unmanned helicopters based on improved Kalman filtering. It optimizes the flight attitude control of the unmanned helicopter under magnetic interference environments through dimensional processing and feedback gain adjustment, ensuring flight stability. The specific implementation method is as follows: The difference assessment process uses a nonlinear feedback adjustment formula, the calculation formula of which is as follows: .

[0072] in, The feedback gain adjustment is dimensionless. The gain scaling factor is dimensionless. The sensitivity coefficient is dimensionless. The magnetic field gradient vector at the current moment is dimensionless. The magnetic field gradient vector from the previous moment is dimensionless. It is a nonlinear suppression function, dimensionless.

[0073] 1. System Initialization and Data Acquisition During the flight of the unmanned helicopter, the flight control system first initializes, acquiring and recording magnetic field and attitude data. The raw data is as follows: Attitude parameter initialization: Roll: 0°, Pitch: 0°, Yaw: 0°.

[0074] Magnetic field data acquisition: The current magnetic field strength is 0.32, 0.52, and 0.75, in tons (T). The previous magnetic field strength was 0.31, 0.5, and 0.77, in tons (T).

[0075] Magnetic field gradient calculation: The magnetic field gradient at the current moment is 0.03, 0.02, -0.02, in units of T / m. The magnetic field gradient at the previous moment was 0.02, 0.01, -0.01, in units of T / m.

[0076] 2. Dimensional processing and dimensionless conversion To meet the dimensionless requirement of the formula, the original data needs to be converted into dimensionless values. The dimensionless conversion process depends on a reference value, usually the sensor's range or maximum value.

[0077] Reference value: Maximum magnetic field strength Maximum magnetic field gradient .

[0078] Dimensionless transformation of magnetic field data:

[0079]

[0080] Dimensionless transformation of magnetic field gradient data:

[0081]

[0082] After dimensionless processing, the units of the magnetic field and magnetic field data are eliminated, and the data becomes dimensionless values.

[0083] 3. Difference Assessment and Gain Calculation The dimensionless magnetic field gradient data is used for difference assessment and to calculate the gain adjustment.

[0084] Calculate the difference in magnetic field gradient:

[0085]

[0086] Applications of nonlinear suppression functions: The gain scaling factor is a coefficient used to control the magnitude of the feedback gain. Adjusting the system's gain scale affects the compensation effect; choosing an appropriate factor is crucial. This value helps avoid overcompensation or slow response. The sensitivity coefficient controls the sensitivity of the feedback gain, representing how sensitive the feedback adjustment is to changes in the magnetic field gradient. It determines how responsive the feedback gain adjustment is to differences in the magnetic field gradient. The gain scaling factor is set. Sensitivity coefficient .

[0087] Calculate the feedback gain adjustment:

[0088] 4. Compensation Feedback Update Results Magnetic field strength fluctuation data: After dimensionless transformation, the difference in magnetic field strength is as follows:

[0089]

[0090] Spatial gradient characteristic information: The difference in magnetic field gradient is 0.1732.

[0091] Feedback gain adjustment: Calculated based on the nonlinear feedback formula. .

[0092] Closed-loop control signal: using dimensionless feedback gain adjustment. Update the parameters of the Kalman filter algorithm to optimize the attitude control signal of the flight control system.

[0093] 5. Data transmission and compensation updates The compensation feedback update results are transmitted to the flight control system through the following two interfaces: UART interface: transmits dimensionless feedback gain adjustment. This is used to update the parameters of the Kalman filter.

[0094] SPI bus: Transmits magnetic field fluctuation data The spatial gradient feature information is 0.1732, which is used for further compensation and adjustment by the flight control system.

[0095] 6. Closed-loop compensation and flight control adjustment The flight control system adjusts the flight attitude and updates the parameters of the Kalman filter based on the compensation feedback. By optimizing flight attitude control through a closed-loop control system, the impact of magnetic interference on flight control is reduced, ensuring flight accuracy and stability.

[0096] Through the above steps, real-time compensation for magnetic interference during the flight of the unmanned helicopter was achieved. By performing dimensionless processing on the magnetic field data and magnetic field gradient data, and combining appropriate gain scaling and sensitivity coefficient settings, the system adjusted the parameters of the Kalman filter to ensure flight attitude stability. Experimental results show that appropriate feedback gain adjustment can significantly reduce the impact of magnetic interference on flight control, improve the system's response speed and control accuracy, and ensure stable flight of the unmanned helicopter in complex magnetic field environments.

[0097] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A real-time magnetic interference compensation method for unmanned helicopters based on improved Kalman filtering, characterized in that, include: S1. The unmanned helicopter flight control system receives a trigger command and processes it to form a compensation start signal. The response processing includes flight control program initialization, attitude parameter loading, and sensor self-test operation. S2. Time synchronization processing is performed on the magnetic field joint sensing data collected during the flight of the unmanned helicopter to form a multi-source flight dataset. The time synchronization processing adopts a unified timestamp calibration mechanism. S3. Perform state estimation processing on the multi-source flight dataset to form magnetic interference estimation results. The state estimation processing adopts an improved Kalman filter algorithm. The improved Kalman filter algorithm adopts a joint state space equation, which includes an unmanned helicopter dynamics model and a magnetic field distribution model. S4. Perform adaptive compensation processing on the magnetic interference estimation results to form magnetic field correction data. The adaptive compensation processing includes adjusting the compensation coefficient and calculating the compensation parameters. S5. Perform difference evaluation processing on the magnetic field correction data to form a compensation feedback update result. The difference evaluation processing includes magnetic field strength fluctuation comparison and gradient feature analysis. Transmit the compensation feedback update result to the flight control system. The compensation feedback update result updates the parameters of the improved Kalman filter algorithm to form a closed-loop compensation mechanism.

2. The real-time magnetic interference compensation method for unmanned helicopters based on improved Kalman filtering according to claim 1, characterized in that: The flight control program initialization includes loading flight control parameters, setting initial values ​​for the attitude control loop PID, and automatic correction of magnetometer bias. The attitude parameter loading includes attitude angle calculation, and the sensor self-test operation includes amplitude and noise characteristic detection.

3. The real-time magnetic interference compensation method for unmanned helicopters based on improved Kalman filtering according to claim 1, characterized in that: The combined magnetic field sensing data includes magnetometer data, accelerometer data, gyroscope data, and GPS data.

4. The real-time magnetic interference compensation method for unmanned helicopters based on improved Kalman filtering according to claim 1, characterized in that: The improved Kalman filter algorithm includes prediction, updating, and adaptive adjustment. The joint state-space equation adopts a dynamic weighting and nonlinear correction coupling formula, the expression of which is: ， in, The result is the state estimation for the next time step. The current state estimation vector, To observe the residuals, This is the dynamic gain matrix, with values ​​ranging from 0 to 1. It is a linear weight matrix. It is a non-linear weight matrix. Let L be the square of the L2 norm of the residual vector. It is a hyperbolic tangent function with an amplitude range of [-1, 1].

5. The real-time magnetic interference compensation method for unmanned helicopters based on improved Kalman filtering according to claim 4, characterized in that: The unmanned helicopter dynamics model outputs the current state estimation vector, which includes attitude angle state variables and angular velocity state variables. The magnetic field distribution model outputs the observation residual, which includes a geomagnetic reference vector, a direction cosine matrix, and magnetic disturbance components. The unmanned helicopter dynamics model and the magnetic field distribution model are bidirectionally coupled.

6. The real-time magnetic interference compensation method for unmanned helicopters based on improved Kalman filtering according to claim 1, characterized in that: The compensation coefficient is dynamically corrected based on the attitude angular velocity, acceleration and magnetic disturbance rate of the unmanned helicopter, and the compensation parameter is the geomagnetic direction vector of the unmanned helicopter dynamic model and the magnetic field distribution model.

7. The real-time magnetic interference compensation method for unmanned helicopters based on improved Kalman filtering according to claim 1, characterized in that: The difference assessment process employs a nonlinear feedback adjustment formula, the calculation formula of which is as follows: ， in, For feedback gain adjustment, This is the gain scaling factor. This is the sensitivity coefficient. The magnetic field gradient vector at the current moment. This represents the magnetic field gradient vector from the previous moment. This is a nonlinear suppression function.

8. The real-time magnetic interference compensation method for unmanned helicopters based on improved Kalman filtering according to claim 1, characterized in that: The compensation feedback update results include magnetic field strength fluctuation data, spatial gradient feature information, parameter update amount, and closed-loop control signal.

9. The real-time magnetic interference compensation method for unmanned helicopters based on improved Kalman filtering according to claim 8, characterized in that: The compensation feedback update result is transmitted using a UART interface and an SPI bus. The UART interface transmits the parameter update amount and the closed-loop control signal, while the SPI bus transmits the magnetic field strength fluctuation data and the spatial gradient characteristic information.

10. A real-time magnetic interference compensation system for unmanned helicopters based on improved Kalman filtering, characterized in that, include: The sensor module collects magnetic field and attitude information during the flight of the unmanned helicopter and outputs joint sensing data of flight attitude and magnetic field. The sensor module includes a three-axis magnetometer, a three-axis accelerometer, a three-axis gyroscope and a GPS module. The flight control main control unit performs time synchronization and state estimation processing on the joint sensing data of flight attitude and magnetic field to form a magnetic interference estimation result. The time synchronization and state estimation processing adopts an improved Kalman filter algorithm. The improved Kalman filter algorithm establishes a joint state space equation based on the dynamic model of the unmanned helicopter and the magnetic field distribution model. The storage unit stores the flight control program, aeromagnetic compensation algorithm, historical compensation parameters, and flight data; A communication unit that transmits the magnetic interference estimation result, the communication unit including a wireless communication module and a data interface; An execution unit is provided, which adjusts the flight attitude of the unmanned helicopter based on the magnetic interference estimation result. The execution unit includes a motor controller and a servo controller.

Images