Novelty reconstruction aided integrated navigation method under denial environment

Abstract

This invention discloses a method for innovation reconstruction-assisted integrated navigation in a denied environment, comprising the following steps: Step 1, information acquisition; Step 2, innovation statistics; Step 3, GNSS signal loss monitoring; Step 4, innovation reconstruction; Step 5, Kalman filter update based on reconstructed innovation; Step 6, switching after GNSS signal recovery. This invention addresses short-term GNSS loss by using reconstructed innovation generated from the mean and variance of historical innovations to replace the missing real innovations and incorporate them into a modified Kalman filter, maintaining the correction of INS navigation parameters. It does not rely on high-precision inertial devices, has low computational load and resource consumption, and is suitable for real-time navigation scenarios such as vehicle and aviation applications. It overcomes the shortcomings of high-precision devices being costly and AI algorithms having poor real-time performance. Furthermore, it does not require modification of existing GNSS receivers or INS hardware, nor does it require deep hardware integration or ultra-tightly coupled complex signal processing, facilitating upgrades and promotion based on existing systems.

Description

Technical Field

[0001] This invention relates to the field of navigation system technology, specifically to an information reconstruction-assisted integrated navigation method in a denied environment. Background Technology

[0002] GNSS / INS integrated navigation systems can leverage the complementary strengths of each system to improve the accuracy and reliability of the navigation system. Existing GNSS / INS integrated navigation systems can be categorized into loosely integrated, tightly integrated, and deeply integrated systems based on their integration methods. Loosely integrated systems use satellite-assisted correction of inertial errors. The observed values ​​are the difference between the position and velocity measured by GNSS and the predicted information by INS. This difference is input into a Kalman filter. It features simple operation, ease of engineering implementation, and navigation redundancy, and can improve positioning accuracy and suppress INS attitude divergence through error compensation. Tightly integrated systems use a combination of satellite and inertial observations, with pseudorange, pseudorange rate, and carrier phase observations as the basic model. It offers high integrated navigation accuracy, dynamism, and robustness. Deeply integrated systems connect satellite tracking signals to the integrated navigation system within a single filter to enhance the ability to track satellite signals. This hardware-level integration method can improve the signal-to-noise ratio of GNSS tracking signals, reduce the impact of multipath effects, and enable rapid acquisition after signal obstruction or interruption. However, this method is still in the research and development stage and has not yet been widely adopted. All three of the above combination methods use GNSS measurement signals as the core input. Therefore, in a denied environment, that is, when GNSS lock is lost and measurement signals are lost, the accuracy of the combined navigation will drop sharply and cannot meet the positioning and navigation requirements.

[0003] To address the GNSS lock-out problem, some technologies employ high-precision inertial devices (INS) to improve the independent operating accuracy of the INS. However, high-precision INS suffers from high R&D costs and difficulty in miniaturization, hindering widespread adoption. Other technologies utilize AI algorithms to assist Kalman filtering, which can compensate for lock-out errors in some scenarios. However, the algorithms have limited adaptability to non-Gaussian noise and strongly nonlinear measurements, and the computational process is complex with insufficient real-time performance. Some technologies employ ultra-tight coupling structures to improve the anti-interference capability of the GNSS / INS integrated navigation system, thereby improving the performance of integrated navigation during short-term GNSS lock-out. However, ultra-tight coupling structures require deep integration of the GNSS tracking loop and the INS, resulting in high structural complexity and difficulty in implementation. Some technologies focus on specific scenarios such as tunnels and underground parking lots, developing targeted navigation products. However, these suffer from poor versatility and cannot flexibly adapt to various navigation scenarios. Summary of the Invention

[0004] The purpose of this invention is to provide an information reconstruction-assisted integrated navigation method in a denied environment to solve the problems mentioned in the background art.

[0005] To achieve the above objectives, the present invention provides the following technical solution: a method for innovation reconstruction-assisted integrated navigation in a denied environment, comprising the following steps: Step 1, information acquisition; Step 2, innovation statistics; Step 3, GNSS signal loss monitoring; Step 4, innovation reconstruction; Step 5, Kalman filter update based on reconstructed innovation; Step 6, switching after GNSS signal recovery;

[0006] In step one above, GNSS measurement information and INS prediction information are obtained during the operation of the GNSS / INS integrated navigation system.

[0007] In step two above, based on the information obtained in step one, the innovation sequence is calculated, and the mean and variance of the innovation sequence are statistically analyzed.

[0008] In step three above, the GNSS signal reception status is monitored in real time to determine whether GNSS signal loss has occurred.

[0009] In step four above, if GNSS signal lock-off is detected, a Gaussian random variable independent and identically distributed with respect to the original innovation is generated based on the mean and variance of the innovation sequence statistically obtained in step two, and this Gaussian random variable is used as the reconstructed innovation.

[0010] In step five above, the reconstructed information generated in step four is substituted into the modified Kalman filter equation to maintain the corrective effect of the Kalman filter on the INS navigation parameters and achieve continuous estimation of the navigation state.

[0011] In step six above, when the GNSS signal is detected to have returned to normal, the generation of reconstructed information is stopped, and the GNSS / INS measurement difference is resumed as the original information input to the Kalman filter.

[0012] Preferably, in step one, the GNSS / INS integrated navigation system adopts a loosely coupled mode, specifically: the position and velocity differences output by the independently operating GNSS and INS are used as the observations of the Kalman filter, and the error estimate output by the Kalman filter is used to correct the INS navigation parameters.

[0013] Preferably, in step two, the information sequence The calculation formula is as follows:

[0014]

[0015] in These are GNSS measurements. For INS The one-step optimal forecast estimate, and .

[0016] Preferably, in step two, the information sequence It has the following properties:

[0017]

[0018] in, Represents the mathematical expectation. Let be the covariance matrix of the new information. For the measurement matrix, The filter covariance matrix is... This is the measurement noise covariance matrix.

[0019] Preferably, in step three, the specific conditions for determining whether GNSS signal loss has occurred include the following two: the number of observable GNSS satellites is less than 4, and the GNSS signal strength is lower than a preset threshold.

[0020] Preferably, in step five, the modified Kalman filter equation specifically includes the following sub-equations:

[0021] State prediction equation:

[0022]

[0023] in for Time-prior state estimation for Post-hoc state estimation at time step This is the state transition matrix;

[0024] Prediction error covariance equation:

[0025]

[0026] in for The prior error covariance matrix at time step, for The posterior error covariance matrix at time step [time]. Here is the state transition matrix. The system noise covariance matrix;

[0027] Kalman gain equation:

[0028]

[0029] in for Moment-time Kalman gain, For the measurement matrix, The measurement noise covariance matrix;

[0030] State update equation:

[0031]

[0032] in for Predicted state value at time of day for Real-time state update value, The reconstructed information generated in step 3;

[0033] Covariance update equation:

[0034]

[0035] in for 3D identity matrix.

[0036] Compared with existing technologies, the beneficial effects of this invention are as follows: This invention addresses short-term GNSS lockout by using reconstructed information generated from the mean and variance of historical information to replace the missing true information and incorporating it into a modified Kalman filter to maintain the correction of INS navigation parameters. It does not rely on high-precision inertial devices, has low computational load and low resource consumption, and is suitable for real-time navigation scenarios such as vehicle and aviation. It overcomes the shortcomings of high-precision devices being costly and AI algorithms having poor real-time performance. At the same time, it does not require modification of existing GNSS receivers or INS hardware, nor does it require deep hardware integration or ultra-tightly coupled complex signal processing, making it easy to upgrade and promote based on existing systems. Attached Figure Description

[0037] Figure 1 This is a flowchart of the method of the present invention;

[0038] Figure 2 A schematic diagram illustrating the geometric meaning of the new information;

[0039] Figure 3 The state estimation diagram based on sampled information reconstruction is shown for a continuous loss of lock-up during the boost phase of 0.1 s; where (a) pitch angle error; (b) yaw angle error; (c) roll angle error; (d) X-axis position error; (e) Y-axis position error; (f) Z-axis position error; (g) X-axis velocity error; (h) Y-axis velocity error; (i) Z-axis velocity error;

[0040] Figure 4 The state estimation diagram based on sampled information reconstruction is given under the condition of continuous loss of lock for 1 second during the boost phase; where (a) pitch angle error; (b) yaw angle error; (c) roll angle error; (d) X-axis position error; (e) Y-axis position error; (f) Z-axis position error; (g) X-axis velocity error; (h) Y-axis velocity error; (i) Z-axis velocity error;

[0041] Figure 5The state estimation diagram based on sampled information reconstruction is given under the condition of continuous loss of lock for 10s during the boost phase; where (a) pitch angle error; (b) yaw angle error; (c) roll angle error; (d) X-axis position error; (e) Y-axis position error; (f) Z-axis position error; (g) X-axis velocity error; (h) Y-axis velocity error; and (i) Z-axis velocity error. Detailed Implementation

[0042] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0043] Please see the appendix Figure 1 -Appendix Figure 2 The present invention provides an embodiment of a novelty reconstruction-assisted integrated navigation method in a denied environment, comprising the following steps: Step 1, information acquisition; Step 2, novelty statistics; Step 3, GNSS signal loss monitoring; Step 4, novelty reconstruction; Step 5, Kalman filter update based on reconstructed novelty; Step 6, switching after GNSS signal recovery.

[0044] In step one above, GNSS measurement information and INS prediction information are obtained during the operation of the GNSS / INS integrated navigation system. The GNSS / INS integrated navigation system adopts a loosely coupled mode, specifically: the position and velocity differences output by the independently operating GNSS and INS are used as the observations of the Kalman filter, and the error estimate output by the Kalman filter is used to correct the INS navigation parameters.

[0045] In step two above, based on the information obtained in step one, the innovation sequence is calculated, and the mean and variance of the innovation sequence are statistically analyzed; wherein, the innovation sequence... The calculation formula is as follows:

[0046]

[0047] in These are GNSS measurements. For INS The one-step optimal forecast estimate, and ;

[0048] New sequence It has the following properties:

[0049]

[0050] in, Represents the mathematical expectation. Let be the covariance matrix of the new information. For the measurement matrix, The filter covariance matrix is... The measurement noise covariance matrix;

[0051] In step three above, the GNSS signal reception status is monitored in real time to determine whether GNSS signal loss has occurred. The specific conditions for determining whether GNSS signal loss has occurred include the following two: the number of observable GNSS satellites is less than 4, and the GNSS signal strength is lower than a preset threshold.

[0052] In step four above, if GNSS signal lock-off is detected, a Gaussian random variable independent and identically distributed with respect to the original innovation is generated based on the mean and variance of the innovation sequence statistically obtained in step two, and this Gaussian random variable is used as the reconstructed innovation.

[0053] In step five above, the reconstructed information generated in step four is substituted into the modified Kalman filter equation to maintain the Kalman filter's corrective effect on the INS navigation parameters, thus achieving continuous estimation of the navigation state. The modified Kalman filter equation specifically includes the following sub-equations:

[0054] State prediction equation:

[0055]

[0056] in for Time-prior state estimation for Post-hoc state estimation at time step This is the state transition matrix;

[0057] Prediction error covariance equation:

[0058]

[0059] in for The prior error covariance matrix at time step, for The posterior error covariance matrix at time step [time]. Here is the state transition matrix. The system noise covariance matrix;

[0060] Kalman gain equation:

[0061]

[0062] in for Moment-time Kalman gain, For the measurement matrix, The measurement noise covariance matrix;

[0063] State update equation:

[0064]

[0065] in for Predicted state value at time of day for Real-time state update value, The reconstructed information generated in step 3;

[0066] Covariance update equation:

[0067]

[0068] in for 3D identity matrix;

[0069] In step six above, when the GNSS signal is detected to have returned to normal, the generation of reconstructed information is stopped, and the GNSS / INS measurement difference is resumed as the original information input to the Kalman filter.

[0070] Experimental example:

[0071] To verify the effectiveness of the method proposed in the embodiments, simulation experiments were conducted, and the initial parameters were selected as follows: Earth radius. Earth's rotational angular velocity Gravitational acceleration The launch azimuth angle is rad, longitude of the launch point is The longitude of the launch point is Launch elevation is ellipsoidal flattening First eccentricity square The constant drift error of the gyroscope is Add table constant drift The system noise variance is: The variance of the observation noise is: The unlocking time was set to 0.1s, 1s, and 10s. The experimental results are attached. Figure 3 -Appendix Figure 5 As shown in the figure, the proposed method can effectively handle the situation of lock loss in different consecutive time periods, and has high estimation accuracy.

[0072] Based on the above, the advantages of this invention are as follows: When used, for short-term GNSS lockout, it generates reconstructed information based on the mean and variance of historical information, replaces the missing real information, and substitutes it into the modified Kalman filter to maintain the correction of INS navigation parameters. It does not rely on expensive and high-precision inertial devices. At the same time, the calculation of this method only involves information statistics and Gaussian variable generation, without the high complexity of AI algorithms. It has low computational load and low resource consumption, and is suitable for real-time navigation scenarios such as vehicle and aviation. It makes up for the shortcomings of high cost of high-precision devices and poor real-time performance of AI algorithms. Furthermore, this method does not require modification of existing GNSS receivers or INS hardware, nor does it require deeply integrated hardware or ultra-tightly coupled complex signal processing logic. It has a simple structure, low technical threshold, and is easy to upgrade and promote based on existing systems.

[0073] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the invention can be implemented in other specific forms without departing from its spirit or essential characteristics. Therefore, the embodiments should be considered in all respects as exemplary and non-limiting, and the scope of the invention is defined by the appended claims rather than the foregoing description. Thus, all variations falling within the meaning and scope of equivalents of the claims are intended to be included within the present invention. No reference numerals in the claims should be construed as limiting the scope of the claims.

Claims

1. A method for information reconstruction-assisted integrated navigation in a denial environment, comprising the following steps: Step 1: Information Acquisition; Step 2: Information Statistics; Step 3, GNSS signal loss monitoring; Step 4, innovation reconstruction; Step 5, Kalman filter update based on reconstructed innovation; Step 6, handover after GNSS signal recovery; characterized in that: In step one above, GNSS measurement information and INS prediction information are obtained during the operation of the GNSS / INS integrated navigation system. In step two above, based on the information obtained in step one, the innovation sequence is calculated, and the mean and variance of the innovation sequence are statistically analyzed. In step three above, the GNSS signal reception status is monitored in real time to determine whether GNSS signal loss has occurred. In step four above, if GNSS signal lock-off is detected, a Gaussian random variable independent and identically distributed with respect to the original innovation is generated based on the mean and variance of the innovation sequence statistically obtained in step two, and this Gaussian random variable is used as the reconstructed innovation. In step five above, the reconstructed information generated in step four is substituted into the modified Kalman filter equation to maintain the corrective effect of the Kalman filter on the INS navigation parameters and achieve continuous estimation of the navigation state. In step six above, when the GNSS signal is detected to have returned to normal, the generation of reconstructed information is stopped, and the GNSS / INS measurement difference is resumed as the original information input to the Kalman filter.

2. The information reconstruction-assisted integrated navigation method under denied conditions according to claim 1, characterized in that: In step one, the GNSS / INS integrated navigation system adopts a loosely coupled mode, specifically: the position and velocity differences output by the independently operating GNSS and INS are used as the observations of the Kalman filter, and the error estimate output by the Kalman filter is used to correct the INS navigation parameters.

3. The information reconstruction-assisted integrated navigation method under denied conditions according to claim 1, characterized in that: In step two, the new information sequence The calculation formula is as follows: ， in These are GNSS measurements. For INS The one-step optimal forecast estimate, and .

4. The information reconstruction-assisted integrated navigation method under denied conditions according to claim 1, characterized in that: In step two, the new information sequence It has the following properties: ， in, Represents the mathematical expectation. Let be the covariance matrix of the new information. For the measurement matrix, Here is the filter covariance matrix. This is the measurement noise covariance matrix.

5. The information reconstruction-assisted integrated navigation method under denied conditions according to claim 1, characterized in that: In step three, the specific conditions for determining whether a GNSS signal loss has occurred include the following two: The number of observable GNSS satellites is less than 4, and the GNSS signal strength is below a preset threshold.

6. The information reconstruction-assisted integrated navigation method under denied conditions according to claim 1, characterized in that: In step five, the modified Kalman filter equation specifically includes the following sub-equations: State prediction equation: ， in for Time-prior state estimation for Post-hoc state estimation at time step This is the state transition matrix; Prediction error covariance equation: ， in for The prior error covariance matrix at time step, for The posterior error covariance matrix at time step [time]. Here is the state transition matrix. Here is the system noise covariance matrix; Kalman gain equation: ， in for Moment-time Kalman gain, For the measurement matrix, The measurement noise covariance matrix; State update equation: ， in for Predicted state value at time of day for Real-time state update value, The reconstructed information generated in step 3; Covariance update equation: ， in for 3D identity matrix.

Images