A Multimodal Data Fusion Method for Hypersonic Vehicle Navigation and Target Relocation

Abstract

This invention discloses a multimodal data fusion method for navigation and target relocation of hypersonic vehicles. Based on a non-equilibrium plasma physics model, the peak electron density and sheath thickness of the plasma sheath are calculated in real time according to flight state parameters. Then, the electromagnetic wave attenuation coefficients of each sensor channel are calculated, and error compensation is performed on the original measurements from the inertial measurement unit, star sensor, infrared imaging sensor, and synthetic aperture radar. Fusion weights are calculated based on the attenuation coefficients and measurement residuals, and the compensated measurements are weighted and fused to output a joint state estimate of the vehicle's attitude and the target's position. When target relocation is triggered, a particle swarm optimization algorithm is used to search for a new target position within a candidate region. This invention solves the problems of decreased navigation accuracy and target loss for hypersonic vehicles in plasma sheath interference environments, and enables continuous tracking of time-sensitive moving targets.

Description

Technical Field

[0001] This invention belongs to the field of aircraft navigation and target tracking technology, specifically relating to a hypersonic aircraft navigation and target relocation method based on multimodal data fusion. Background Technology

[0002] During high-speed flight, near-space hypersonic vehicles experience high temperatures due to intense friction with the atmosphere, causing the gas on the vehicle's surface to ionize and form a plasma sheath that envelops the vehicle. This plasma sheath has a high electron density, which strongly absorbs, reflects, and scatters electromagnetic waves, resulting in severe attenuation or even complete interruption of communication and navigation signals between the vehicle and the outside world—a phenomenon known as the "blackout."

[0003] In missions involving continuous tracking of time-sensitive moving targets, the blackout problem presents a significant challenge. On one hand, traditional navigation methods relying on external signals, such as satellite navigation and radio measurement, experience a sharp decline in performance under plasma sheath interference, making it difficult to guarantee the accuracy of the aircraft's attitude estimation. On the other hand, sensors used for target detection, such as infrared imaging and synthetic aperture radar, are also affected by plasma attenuation, leading to a deterioration in the quality of target observation data and a high risk of target loss. Current technologies primarily employ inertial navigation systems for autonomous navigation in plasma sheath environments, but their errors accumulate over time. For multi-sensor fusion, fixed-weight or covariance-based fusion methods are typically used, failing to fully consider the differentiated impact of plasma on sensors at different frequency bands. Furthermore, there is a lack of effective target relocation mechanisms to address target loss.

[0004] Therefore, there is an urgent need for a method that can achieve high-precision navigation and continuous target tracking in a plasma sheath interference environment. Summary of the Invention

[0005] The purpose of this invention is to address the problems of decreased navigation accuracy and target loss faced by near-space hypersonic vehicles in plasma sheath interference environments. It provides a multi-modal data fusion method for hypersonic vehicle navigation and target relocation, which predicts sheath parameters in real time by establishing a plasma physics model, performs plasma attenuation compensation and adaptive weighted fusion on multi-source sensor measurements, and uses a particle swarm optimization algorithm to achieve rapid target relocation when the target is lost, thereby ensuring the vehicle's continuous tracking capability of time-sensitive moving targets in blackout environments.

[0006] The specific technical solution is as follows:

[0007] A multimodal data fusion method for navigation and target relocation of hypersonic vehicles is applied to a scenario where a near-space hypersonic glide vehicle continuously tracks a time-sensitive moving target in a plasma sheath interference environment. The method includes the following steps:

[0008] Step 1: Based on the current flight status parameters of the aircraft, calculate the peak electron density and sheath thickness of the plasma sheath in real time using a non-equilibrium plasma physics model.

[0009] Step 2: Based on the peak electron density and thickness of the plasma sheath, calculate the electromagnetic wave attenuation coefficient of each sensor channel, and then perform error compensation on the original measurement values ​​of the inertial measurement unit, star sensor, infrared imaging sensor and synthetic aperture radar to obtain the compensated measurement values.

[0010] Step 3: Calculate the fusion weight of each sensor based on the attenuation coefficient and the measurement residual after compensation, perform weighted fusion on the compensated measurement values, and output the fused state estimate of the aircraft pose and the target position.

[0011] Step 4: When target relocation is triggered, based on the joint state vector and target kinematic constraints, the particle swarm optimization algorithm is used to search for a new target position in the candidate region to complete the target relocation.

[0012] Furthermore, the method for real-time calculation of the peak electron density and sheath thickness of the plasma sheath based on a non-equilibrium plasma physics model is as follows:

[0013] Get the current Mach number of the aircraft Atmospheric density corresponding to flight altitude and station temperature Electron density peak : ;in, These are calibration coefficients related to the aircraft's shape. The ionization energy of air molecules, Boltzmann constant; sheath thickness : ;in, This is the geometric correction factor. The characteristic radius of the aircraft's nose. For reference atmospheric density.

[0014] Furthermore, the electromagnetic wave attenuation coefficient of each sensor channel was calculated separately. The measurement error compensation method is based on the peak electron density. and sheath thickness Calculate the characteristic frequency of plasma : ;in, The amount of electron charge. The vacuum permittivity, For electronic quality; for operating frequency of The Each sensor has an electromagnetic wave attenuation coefficient. The calculation is as follows: ;in, For electron collision frequency, It is the speed of light.

[0015] when When a sensor channel is determined to be completely interrupted, the fusion weight of that sensor will be... .

[0016] Furthermore, when the attenuation coefficient of a sensor exceeds a critical threshold... At that time, that is The confidence factor of the sensor is forced to zero, and the critical threshold is... The fusion weights are set separately according to the sensor type; correspondingly, the fusion weights of the sensor are... .

[0017] Furthermore, error compensation is performed on the raw measurements from the inertial measurement unit, star sensor, infrared imaging sensor, and synthetic aperture radar to obtain the compensated measurements. The method for obtaining the compensated measurements is as follows: From the original measurement value The result of the correction is: ;in, This is a system bias correction term related to the attenuation coefficient, obtained from offline calibration.

[0018] Furthermore, the fusion weights The calculation method is as follows: Define the first Reliability factor of each sensor : ; To compensate for the measurement residuals: ;in, For the first The observation function of each sensor, Measure its noise covariance matrix. As an observation dimension, and It is an adjustable weighting factor.

[0019] Fusion weights The following is obtained by normalizing the credibility factor: ;in, The total number of sensors participating in the fusion. Fusion state estimation Solve using the weighted least squares method:

[0020] State vector It includes the aircraft's position, speed, attitude, and the target's relative position and speed.

[0021] Furthermore, the target relocation triggering conditions and search methods are as follows: Define the relocation triggering criteria. : ;in, For the first The sequence of information at any given moment express Fusion measurements from time sensors The equivalent observation function after fusion. For based on Time information Prediction and estimation of state at time step. The new information covariance matrix, K The sliding window length represents the number of historical moments used to calculate the criterion.

[0022] when Target relocation is triggered at certain times, where The preset threshold is used; target relocalization employs a particle swarm optimization algorithm, based on fused state estimation. The target position component is centered at a radius of . The particle swarm is initialized within a spherical region, where For the target maximum speed estimation, The duration of target loss.

[0023] The particle fitness function is defined as: ;in, In particle state, For the target kinematic constraint penalty term, To constrain the weights; once the particle swarm converges, the globally optimal particle position is used as the relocation result.

[0024] Furthermore, the peak electron density and sheath thickness Correction is achieved through online recursive identification: the measured attenuation value of the received signal is obtained using an airborne multi-frequency beacon. Compared with the model predicted decay value To determine the deviation, construct the parameter correction equation:

[0025] ;in, and The corrections for electron density and sheath thickness are respectively estimated using the recursive least squares method: ;in, For Jacobian matrices, This is the recursive gain matrix.

[0026] Corrected electron density peak Corrected sheath thickness Update the attenuation coefficient again using the corrected parameters. .

[0027] Furthermore, when the attenuation coefficient of a certain sensor recovers to At that time, it is reintegrated into the fusion loop and triggers target relocation, where To prevent hysteresis caused by frequent switching.

[0028] Compared with the prior art, the beneficial effects of this invention are:

[0029] This invention can dynamically predict the peak electron density and sheath thickness based on flight status, providing an accurate basis for sensor error compensation; it designs a two-factor reliability evaluation mechanism based on plasma attenuation coefficient and measurement residual, realizing adaptive weighted fusion of multi-sensor data, effectively suppressing the negative impact of interfered sensors on the fusion results; it proposes a target relocation trigger criterion based on innovation sequence statistical test and a target search algorithm based on particle swarm optimization, which can quickly recover tracking after target loss; it introduces an online recursive identification mechanism for plasma parameters, using measured attenuation data to correct model parameters, improving the model's adaptability under complex flight conditions. Attached Figure Description

[0030] Figure 1 This is a flowchart of a hypersonic vehicle navigation and target relocation method based on multimodal data fusion according to the present invention. Detailed Implementation

[0031] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of this invention are described clearly and completely below. Obviously, the described embodiments are only a part of the embodiments of this invention, not all of them. Based on the embodiments of this invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this invention.

[0032] It should be noted that the method of the present invention is applied to the scenario of continuous tracking of time-sensitive moving targets by a near-space hypersonic glide vehicle in a plasma sheath interference environment.

[0033] This embodiment uses a typical near-space hypersonic glider performing a time-sensitive ground target tracking mission as its application background. The glider glides at 5-20 times the speed of sound within an altitude range of 20-60 km, and needs to continuously track ground vehicles during flight. The sensor system onboard the glider includes: an inertial measurement unit (IMU) for measuring the glider's own acceleration and angular velocity; a star sensor for providing attitude reference; an infrared imaging sensor for detecting the infrared radiation characteristics of the target; and an X-band synthetic aperture radar (SAR) for acquiring high-resolution images of the target.

[0034] Under hypersonic flight conditions, the temperature at the stagnation point of the aircraft's nose can reach thousands of Kelvin, causing thermal ionization of the surrounding air and forming a plasma sheath. The electron density of the plasma sheath varies with the flight Mach number and altitude, typically ranging from 10^16 to 10^19 m^-3, corresponding to plasma characteristic frequencies between 1 and 30 GHz. When the sensor's operating frequency is lower than the plasma characteristic frequency, electromagnetic waves cannot penetrate the sheath, causing the sensor to completely fail; when the operating frequency is higher than the characteristic frequency but not significantly different, the electromagnetic waves are significantly attenuated, resulting in a severe decrease in measurement accuracy.

[0035] In this embodiment, the typical operating frequencies of the four sensors are as follows: the IMU does not rely on electromagnetic wave transmission and is not directly affected by plasma but has drift error; the star sensor operates in the visible / near-infrared band (frequency about 10^14 Hz) and is less affected by plasma; the infrared imaging sensor operates in the mid-far-infrared band (frequency about 10^13 Hz) and is affected by plasma scattering; the SAR operates in the X-band (about 10 GHz) and may be completely blocked in a strong plasma environment.

[0036] Example

[0037] like Figure 1 The diagram shown is a flowchart of a hypersonic vehicle navigation and target relocation method based on multimodal data fusion according to the present invention, which includes the following steps:

[0038] Step 1: Based on the current flight status parameters of the aircraft, calculate the peak electron density and sheath thickness of the plasma sheath in real time using a non-equilibrium plasma physics model.

[0039] The method for real-time calculation of the peak electron density and sheath thickness of the plasma sheath based on a non-equilibrium plasma physics model is as follows: obtain the current Mach number of the spacecraft. Atmospheric density corresponding to flight altitude and station temperature .

[0040] peak electron density : ;in, These are calibration coefficients related to the aircraft's shape. The ionization energy of air molecules, is the Boltzmann constant.

[0041] Sheath thickness : ;in, This is the geometric correction factor. The characteristic radius of the aircraft's nose. For reference atmospheric density.

[0042] Taking an aircraft flying at an altitude of 40km and a Mach number of 10 as an example, the atmospheric density at this time is approximately The stagnation temperature is approximately 2500K. Assume the characteristic radius of the aircraft's nose... =0.3m, calibration coefficient Geometric correction factor =1.2, with reference atmospheric density taken as the standard atmospheric density at sea level. air molecule ionization energy Boltzmann constant J / K.

[0043] Substituting into the formula for calculating the peak electron density, we can obtain... Substituting into the formula for calculating sheath thickness, we can obtain... ≈3.7cm; The above calculation results show that under typical flight conditions of Mach 10 and an altitude of 40km, the peak electron density of the plasma sheath is approximately The sheath thickness is approximately 3.7 cm. As the Mach number increases or the flight altitude decreases, the electron density will further increase, and the sheath's attenuation effect on electromagnetic waves will become more significant.

[0044] Step 2: Based on the peak electron density and thickness of the plasma sheath, calculate the electromagnetic wave attenuation coefficient of each sensor channel, and then perform error compensation on the original measurement values ​​of the inertial measurement unit, star sensor, infrared imaging sensor and synthetic aperture radar to obtain the compensated measurement values.

[0045] Calculate the electromagnetic wave attenuation coefficient for each sensor channel. The measurement error compensation method is as follows:

[0046] Based on electron density peak and sheath thickness Calculate the characteristic frequency of plasma : ;in, The amount of electron charge. The vacuum permittivity, For electronic quality; for operating frequency of The Each sensor has an electromagnetic wave attenuation coefficient. The calculation is as follows: ;in, For electron collision frequency, It is the speed of light.

[0047] Known electron charge C, vacuum permittivity F / m, electron mass kg, substituting into the formula, we get ≈5.1GHz.

[0048] For X-band SAR (operating frequency) =10GHz), due to Electromagnetic waves can penetrate the plasma sheath, but are attenuated. Assuming the electron collision frequency... Hz, speed of light m / s, substituting into the attenuation coefficient formula, we can obtain ≈0.85 indicates that the SAR signal power is attenuated by about 57% after passing through the plasma sheath.

[0049] For star sensors (operating frequency approximately...) Hz) and infrared imaging sensor (operating frequency approximately (Hz), because its operating frequency is much higher than the characteristic frequency of plasma. =5.1GHz, the calculated attenuation coefficient Very small (approximately) The plasma sheath (on the order of magnitude) can be approximated as unaffected by plasma decay; however, the high-temperature radiation generated by the plasma sheath will cause background interference to the infrared sensor, which needs to be suppressed through subsequent signal processing.

[0050] For inertial measurement units (IMUs), plasma attenuation coefficients are not applicable because they do not rely on electromagnetic waves for measurement. However, IMUs have inherent drift errors, which can be represented by gyroscope bias, accelerometer bias, and corresponding random walks. These errors accumulate over time and need to be corrected through fusion with other sensors.

[0051] when When a sensor channel is determined to be completely interrupted, the fusion weight of that sensor will be... .

[0052] When the attenuation coefficient of a sensor exceeds a critical threshold At that time, that is The confidence factor of the sensor is forced to zero, and the critical threshold is... The fusion weights are set separately according to the sensor type; correspondingly, the fusion weights of the sensor are... .

[0053] When the attenuation coefficient of a certain sensor recovers to At that time, it is reintegrated into the fusion loop and triggers target relocation, where To prevent hysteresis caused by frequent switching.

[0054] The critical attenuation threshold for each sensor is set separately based on its measurement characteristics: SAR critical threshold =2.0 (corresponding to 99% power attenuation); beyond this threshold, SAR measurement data is unreliable; critical threshold for infrared sensors. =1.5, mainly considering interference from plasma thermal radiation background; critical threshold of star sensor =0.5, mainly considering the impact of plasma luminescence on star point recognition. Hysteresis The value is uniformly set to 0.2 to prevent the sensor from frequently entering and exiting the fusion loop near the threshold.

[0055] Taking SAR as an example, when the attenuation coefficient At that time, the original measurement value Magnification compensation is required; the compensated measurement value is... Among them, the systematic bias correction term The results, obtained through ground calibration tests, reflect the systematic biases in SAR ranging and angle measurement under plasma conditions.

[0056] The method for error compensation of the raw measurements from the inertial measurement unit, star sensor, infrared imaging sensor, and synthetic aperture radar to obtain the compensated measurements is as follows: [Compensated Measurement Values] From the original measurement value The result of the correction is: ;in, This is a system bias correction term related to the attenuation coefficient, obtained from offline calibration.

[0057] Step 3: Calculate the fusion weight of each sensor based on the attenuation coefficient and the measurement residual after compensation, perform weighted fusion on the compensated measurement values, and output the fused state estimate of the aircraft pose and the target position.

[0058] The fusion weight The calculation method is as follows: Define the first Reliability factor of each sensor : ; To compensate for the measurement residuals: ;in, For the first The observation function of each sensor, Measure its noise covariance matrix. As an observation dimension, and It is an adjustable weighting factor;

[0059] Fusion weights The following is obtained by normalizing the credibility factor: ;in, The total number of sensors participating in the fusion. Fusion state estimation Solve using the weighted least squares method:

[0060] State vector It includes the aircraft's position, speed, attitude, and the target's relative position and speed.

[0061] Weighting factors and The values ​​are respectively and ; The influence of the attenuation coefficient on reliability is controlled; the larger the value, the heavier the penalty for high-attenuation sensors. The degree to which the control measurement residual affects the reliability is determined; the larger the value, the heavier the penalty for abnormal sensors due to residuals.

[0062] Taking the sensor state at a certain moment as an example, assuming the SAR attenuation coefficient at this moment... Normalized residuals The infrared sensor has an attenuation coefficient of 0.1 and a normalized residual of 0.8; the star sensor has an attenuation coefficient... The normalized residual is 0.5; the IMU equivalent attenuation coefficient is... (Unaffected by plasma), the normalized residual is 2.0 (due to the large residual caused by drift accumulation).

[0063] Substitute into the credibility factor formula to calculate: ; ; ; .

[0064] The fusion weights are obtained after normalization: , , , It can be seen that the SAR sensor, which is more affected by plasma attenuation, has the lowest weight, while the star sensor, which has low attenuation and normal residual, has the highest weight.

[0065] Step 4: When target relocation is triggered, based on the joint state vector and target kinematic constraints, the particle swarm optimization algorithm is used to search for a new target position in the candidate region to complete the target relocation.

[0066] The target relocation triggering conditions and search methods are defined as follows: Relocation triggering criteria are defined. : ;in, For the first The sequence of information at any given moment express Fusion measurements from time sensors The equivalent observation function after fusion. For based on Time information Prediction and estimation of state at time step. The new information covariance matrix, The sliding window length represents the number of historical moments used to calculate the criterion.

[0067] Specifically, the state vector The dimension is 15, including: the three-dimensional position of the spacecraft in the geocentric-geocentric coordinate system. 3D velocity Attitude represented by quaternions (simplified to pitch angles in three-dimensional Euler angles) Yaw angle Roll angle ), and the three-dimensional position of the target relative to the aircraft. and three-dimensional velocity The fusion state estimation is updated at a frequency of 100Hz, consistent with the IMU's output frequency.

[0068] when Target relocation is triggered at certain times, where The preset threshold is used; target relocalization employs a particle swarm optimization algorithm, based on fused state estimation. The target position component is centered at a radius of . The particle swarm is initialized within a spherical region, where For the target maximum speed estimation, The duration of target loss.

[0069] The sliding window length is set to 10, meaning the relocation criterion D is calculated using the information sequence from the most recent 10 sampling times. Relocation trigger threshold. Based on the chi-square distribution specification, at a 95% confidence level, for the case with an observation dimension of 3... ≈7.81. When the information sequence at multiple consecutive time points deviates significantly from the zero-mean assumption, criterion D will exceed the threshold, indicating that there is a systematic deviation between the current target state estimate and the actual observation, and relocation needs to be triggered.

[0070] The scenarios that trigger relocation include: (1) the target undergoes a large maneuver, such as a vehicle suddenly turning or accelerating, which leads to an increase in the prediction deviation based on the uniform motion model; (2) the sensor recovers from a completely interrupted state, such as the SAR regaining the target echo after passing through a dense plasma area, but at this time the actual position of the target has deviated significantly from the predicted position; (3) the target reappears after being blocked, and the infrared or SAR sensor re-acquires the target.

[0071] The particle fitness function is defined as:

[0072] ;

[0073] in, In particle state, For the target kinematic constraint penalty term, To constrain the weights; once the particle swarm converges, the globally optimal particle position is used as the relocation result.

[0074] Taking a ground vehicle target as an example to illustrate the determination of the search area, assuming the target's maximum speed is estimated... =30m / s (equivalent to a vehicle traveling at approximately 108km / h), target loss duration =5s, then the search radius =30×5=150m. The particle swarm is initialized within a spherical region with a radius of 150m centered on the last reliable estimated position.

[0075] In this embodiment, the particle swarm size is set to 50 particles, and the maximum number of iterations is 30. The inertia weight of the particle swarm optimization is linearly decreased from 0.9 to 0.4, and the individual learning factor... =2.0, Social Learning Factor =2.0. The target kinematic constraint penalty term g(x_p) considers the following constraints: (1) the target speed does not exceed the maximum speed limit; (2) the target acceleration does not exceed the physically feasible range (for ground vehicles, the lateral acceleration is usually no more than 0.5g); (3) the target position should be within a reasonable terrain elevation range. The constraint weight γ is adaptively adjusted according to the degree of constraint violation, with a typical initial value of 100.

[0076] The convergence criterion for particle swarm optimization is that the change in the global optimal fitness value over five consecutive iterations is less than 1%. The maximum number of iterations may be reached. After convergence, the position of the globally optimal particle is the relocation result. This position will be used as the initial value of the target state for subsequent fusion filtering. At the same time, the covariance matrix of the target state will be reset to a larger initial value to reflect the uncertainty of the state estimation after relocation.

[0077] Furthermore, the peak electron density and sheath thickness Correction is performed using online recursive identification:

[0078] Measured attenuation values ​​of signals received using airborne multi-frequency beacons Compared with the model predicted decay value To determine the deviation, construct the parameter correction equation:

[0079] ;in, and The corrections for electron density and sheath thickness are respectively estimated using the recursive least squares method:

[0080] ;in, For Jacobian matrices, This is the recursive gain matrix;

[0081] Corrected electron density peak Corrected sheath thickness Update the attenuation coefficient again using the corrected parameters. .

[0082] In this embodiment, the airborne multi-frequency beacon system includes beacon receivers for three frequency bands: L-band (1.5 GHz), S-band (3 GHz), and C-band (6 GHz). By receiving known power signals transmitted from ground or space beacon stations and measuring the actual signal power after passing through a plasma sheath, the measured attenuation values ​​for each frequency band are obtained. Because the attenuation characteristics of the three frequencies are different, an overdetermined system of equations can be constructed to improve the robustness of parameter identification.

[0083] The forgetting factor for the recursive least squares method is set to 0.98, enabling the algorithm to adapt to dynamic changes in plasma parameters while also providing some smoothing effect on measurement noise. Parameter correction amount and The initial value is set to 0, and the initial covariance matrix is ​​set to a diagonal matrix, with the larger diagonal elements (e.g., ...). To accelerate the initial convergence speed. When the identified correction exceeds the physically reasonable range (e.g.) or If the current measurement is abnormal, no parameter update will be performed.

[0084] The specific embodiments described above further illustrate the purpose, technical solution, and beneficial effects of the present invention. It should be understood that the above description is only a specific embodiment of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A multimodal data fusion method for navigation and target relocation of hypersonic vehicles, applied to a near-space hypersonic glider continuously tracking time-sensitive moving targets in a plasma sheath interference environment, characterized in that... The method includes the following steps: Step 1: Based on the current flight status parameters of the aircraft, calculate the peak electron density and sheath thickness of the plasma sheath in real time using a non-equilibrium plasma physics model. Step 2: Based on the peak electron density and thickness of the plasma sheath, calculate the electromagnetic wave attenuation coefficient of each sensor channel, and then perform error compensation on the original measurement values ​​of the inertial measurement unit, star sensor, infrared imaging sensor and synthetic aperture radar to obtain the compensated measurement values. Step 3: Calculate the fusion weight of each sensor based on the attenuation coefficient and the measurement residual after compensation, perform weighted fusion on the compensated measurement values, and output the fusion state estimate of the aircraft pose and the target position. Step 4: When target relocation is triggered, based on the joint state vector and target kinematic constraints, the particle swarm optimization algorithm is used to search for a new target position in the candidate region to complete the target relocation.

2. The method according to claim 1, characterized in that, The method for real-time calculation of the peak electron density and sheath thickness of the plasma sheath based on a non-equilibrium plasma physics model is as follows: Get the current Mach number of the aircraft Atmospheric density corresponding to flight altitude and station temperature ; peak electron density : ;in, These are calibration coefficients related to the aircraft's shape. The ionization energy of air molecules, Boltzmann's constant; Sheath thickness : ;in, This is the geometric correction factor. The characteristic radius of the aircraft's nose. For reference atmospheric density.

3. The method according to claim 2, characterized in that, Calculate the electromagnetic wave attenuation coefficient for each sensor channel. The measurement error compensation method is as follows: Based on electron density peak and sheath thickness Calculate the characteristic frequency of plasma : ;in, The amount of electron charge. The vacuum permittivity, For electronic quality; for operating frequency of The Each sensor has an electromagnetic wave attenuation coefficient. The calculation is as follows: ;in, For electron collision frequency, The speed of light; when When a sensor channel is determined to be completely interrupted, the fusion weight of that sensor will be... .

4. The method according to claim 3, characterized in that, When the attenuation coefficient of a sensor exceeds a critical threshold At that time, that is The confidence factor of the sensor is forced to zero, and the critical threshold is... The fusion weights are set separately according to the sensor type; correspondingly, the fusion weights of the sensor are... .

5. The method according to claim 4, characterized in that, The method for error compensation of the raw measurements from the inertial measurement unit, star sensor, infrared imaging sensor, and synthetic aperture radar to obtain the compensated measurements is as follows: [Compensated Measurement Values] From the original measurement value The result of the correction is: ;in, This is a system bias correction term related to the attenuation coefficient, obtained from offline calibration.

6. The method according to claim 5, characterized in that, The fusion weight The calculation method is as follows: Define the first Reliability factor of each sensor : ; To compensate for the measurement residuals: ;in, For the first The observation function of each sensor, Measure its noise covariance matrix. As an observation dimension, and It is an adjustable weighting factor; Fusion weights The following is obtained by normalizing the credibility factor: ;in, The total number of sensors participating in the fusion. Fusion state estimation Solve using the weighted least squares method: State vector It includes the aircraft's position, speed, attitude, and the target's relative position and speed.

7. The method according to claim 4, characterized in that, The target relocation triggering conditions and search methods are defined as follows: Relocation triggering criteria are defined. : ;in, For the first The sequence of information at any given moment express Fusion measurements from time sensors The equivalent observation function after fusion. Based on Time information Prediction and estimation of state at time step. The new information covariance matrix, K The sliding window length represents the number of historical moments used to calculate the criterion. when Target relocation is triggered at certain times, where The preset threshold is used; target relocalization employs a particle swarm optimization algorithm, based on fused state estimation. The target position component is centered at a radius of . The particle swarm is initialized within a spherical region, where For the target maximum speed estimation, For the duration of target loss; The particle fitness function is defined as: ； in, In particle state, For the target kinematic constraint penalty term, To constrain the weights; once the particle swarm converges, the globally optimal particle position is used as the relocation result.

8. The method according to claim 7, characterized in that, The peak electron density and sheath thickness Correction is performed using online recursive identification: Measured attenuation values ​​of signals received using airborne multi-frequency beacons Compared with the model's predicted decay value To determine the deviation, construct the parameter correction equation: ;in, and The corrections for electron density and sheath thickness are respectively estimated using the recursive least squares method: ;in, For Jacobian matrices, This is the recursive gain matrix; Corrected electron density peak Corrected sheath thickness Update the attenuation coefficient again using the corrected parameters. .

9. The method according to claim 8, characterized in that, When the attenuation coefficient of a certain sensor recovers to At that time, it is reintegrated into the fusion loop and triggers target relocation, where To prevent hysteresis caused by frequent switching.

Images