A method for suppressing multi-inertial navigation positioning errors based on Kalman filter identification
Through the multi-inertial navigation positioning error suppression method based on Kalman filtering, the longitude divergence term error in the multi-inertial navigation system is identified and suppressed, and the problem of limited improvement in positioning accuracy in the prior art is solved, and a higher multi-inertial navigation configuration positioning is achieved Accuracy.
Patent Information
- Application Number
- CN202310789256.4
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2023-06-30
- Publication Date
- 2024-11-05
- Estimated Expiration
- 2043-06-30
AI Technical Summary
The existing multi-inertial navigation system cannot effectively suppress longitude divergence term errors, resulting in limited improvement in positioning accuracy.
The multi-inertial guide positioning error suppression method based on Kalman filtering is used to reversely identify other inertial guide positioning characteristics, especially the differences in longitude divergence characteristics, to achieve suppression of longitude divergence term errors by analyzing the estimation results of the multi-inertial guide Kalman filter. Specific steps include time synchronization, establishing a state space model, calculating the gyroscope drift and accelerometer zero-bias estimates, calculating the longitude divergence direction and amplitude parameters, and performing fusion and weighted average of longitude information.
The positioning accuracy of multi-inertial navigation configuration is improved, and the positioning accuracy of the carrier and the comprehensive benefits of the navigation system are improved by identifying and suppressing longitude divergence errors.
Smart Images

Figure SMS_2 
Figure SMS_8 
Figure SMS_21
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of multi-inertial navigation information fusion, and in particular to a multi-inertial navigation positioning error suppression method based on Kalman filter identification. Background Art
[0002] With the continuous optimization and upgrading of the weight and volume of inertial navigation systems, in order to ensure the safety and reliability of inertial navigation means, more and more carriers are choosing three or more multi-inertial navigation configuration schemes. Therefore, it is urgent to tap the potential of multi-inertial navigation configuration and further improve the carrier positioning accuracy based on the performance of single inertial navigation, thereby further improving the comprehensive benefits of multi-inertial navigation configuration schemes.
[0003] According to the working principle of inertial navigation, the factor that has the greatest impact on positioning accuracy is the error of the longitude divergence term over time. This error is stimulated by gyro drift and accumulates over time, seriously affecting the inertial navigation positioning accuracy. In addition, due to the randomness of gyro drift, the longitude divergence characteristics of different inertial navigation systems and the same inertial navigation system under different starting conditions are different. This difference is reflected in the error divergence amplitude on the one hand, and in the divergence direction on the other hand, that is, the two situations of eastward deviation or westward deviation, which brings great difficulty to the suppression of the longitude divergence error.
[0004] Currently, there is no technical solution that can suppress the longitude divergence error, so the positioning accuracy of the existing multi-INS configuration navigation system has been limited. Summary of the invention
[0005] In view of this, the present invention provides a multi-INS positioning error suppression method based on Kalman filter identification, which can reversely identify the differences in other INS positioning characteristics, especially the longitude divergence characteristics, by analyzing the multi-INS Kalman filter estimation results, laying the foundation for longitude divergence error suppression, thereby achieving the purpose of improving the positioning accuracy of the multi-INS configuration.
[0006] To achieve the above object, the present invention provides a multi-INS positioning error suppression method based on Kalman filter identification, which is used to suppress positioning errors in a multi-INS navigation system equipped with more than three INS, and specifically includes the following steps:
[0007] Step 1: Sort more than three INS and name them INS 1, 2, ... N in sequence, and output time synchronization information to each INS through the same time reference device. The time synchronization information includes TOD and second pulse. Each INS uses an external time reference to complete time synchronization and mark the timestamp information in the output longitude and latitude information.
[0008] Step 2: After the normal operation of multiple inertial navigation systems for a set period of time, select one inertial navigation system as the reference inertial navigation system and take the other inertial navigation systems as the reference inertial navigation systems. Perform the following steps:
[0009] Step three, establish the state space model of the reference inertial navigation and the reference inertial navigation; the state space model adopts the inertial navigation position, velocity, and attitude error equations, and uses Kalman filtering to obtain the gyro drift estimate and accelerometer zero bias estimate of the reference inertial navigation at the kth minute when each reference inertial navigation is observed.
[0010] Step 4: Based on the gyro drift estimate and accelerometer bias estimate of the reference inertial navigation system at the kth minute under each reference inertial navigation system observation condition obtained in step 3, calculate the longitude divergence direction parameters between the reference inertial navigation systems.
[0011] Step 5: If the longitude divergence direction parameters of the two reference inertial navigation systems at the kth minute are greater than or equal to 0, the longitude divergence directions of the two reference inertial navigation systems are the same, and it is necessary to further calculate the longitude divergence amplitude parameters between the two reference inertial navigation systems; then, the longitude divergence direction parameters and the longitude divergence amplitude parameters are combined to fuse the longitudes of the two reference inertial navigation systems to obtain the fused longitude information for output.
[0012] Step 6: Switch the reference INS to the next INS, and use the other INS other than the reference INS as the reference INS. Repeat steps 3 to 5 until the fused longitude information between all INS in the multi-INS navigation system is determined. Select two groups of INS with different longitude divergence directions and the smallest divergence amplitude in the same direction, and take the weighted average of their output longitudes as the multi-INS fusion longitude output.
[0013] Furthermore, when the multi-inertial navigation system works normally, it outputs longitude and latitude information, eastward speed, northward speed and vertical speed information, and heading, roll and pitch attitude information.
[0014] Furthermore, in the state-space model, the model observations are the position difference and velocity difference between the reference inertial navigation and the reference inertial navigation; the model state quantities are the position difference, velocity difference, attitude difference between the reference inertial navigation and the reference inertial navigation, and the gyro drift and accelerometer zero bias of the reference inertial navigation.
[0015] Furthermore, the longitude divergence direction parameters between the two reference inertial navigations are calculated, specifically:
[0016] The order of the two reference inertial navigations is p and q respectively, then the longitude divergence direction parameter between the two at the kth minute is , specifically:
[0017]
[0018] in are the estimated gyro drift values of the inertial navigation system p on the x-axis, y-axis and z-axis in the north celestial east coordinate system, are the estimated accelerometer bias values of the inertial navigation system p on the x-axis, y-axis, and z-axis in the north celestial east coordinate system, are the estimated gyro drift values of the inertial navigation q on the x-axis, y-axis and z-axis in the north celestial east coordinate system, are the estimated accelerometer bias values of the inertial navigation system q on the x-axis, y-axis, and z-axis in the north celestial east coordinate system, respectively. The ⊕ operation is defined as: a and b are the objects before and after the symbol ⊕ is processed.
[0019] Furthermore, the longitude divergence amplitude parameters between the two reference inertial navigations are calculated:
[0020]
[0021] The ⊙ operation is defined as: a' and b' are the objects before and after the symbol ⊙ processing respectively.
[0022] Furthermore, the longitude divergence direction parameter and the longitude divergence amplitude parameter are combined to fuse the longitudes of the two reference inertial navigation systems to obtain the fused longitude information for output, specifically:
[0023] like Then the longitude divergence directions of INS p and INS q are the same, and the longitude divergence amplitude of INS p is smaller than that of INS q. When multiple INS are fused, INS p longitude is selected as the fused longitude information output;
[0024] like Then the longitude divergence directions of INS p and INS q are the same, and the longitude divergence amplitude of INS p is greater than that of INS q. When multiple INS are fused, INS q longitude is selected as the fused longitude information output;
[0025] like The longitude divergence directions of INS p and INS q are different. When multiple INS are fused, the weighted average of the longitudes of the two INS is selected as the fused longitude information output.
[0026] Beneficial effects:
[0027] 1. The present invention provides a multi-INS positioning error suppression method based on Kalman filter identification. In the case of three or more multi-INS configurations, the multi-INS output position difference and speed difference are used as observation quantities, and the same state space model and Kalman filter parameters are adopted to estimate the reference INS gyro drift. The multi-INS position information difference as the observation quantity is reversely identified according to the estimation effect, especially the identification of different INS longitude divergence directions. If they are different directions, the multi-INS longitude weighted average is used to achieve longitude divergence error suppression. If they are the same direction, the one with the smallest divergence amplitude is selected as the multi-INS fusion longitude output according to the identification result, which can also achieve longitude divergence error suppression. Thereby achieving the purpose of improving the positioning accuracy of the multi-INS configuration. DETAILED DESCRIPTION
[0028] The present invention is described in detail below with reference to the following embodiments.
[0029] The invention provides a multi-INS positioning error suppression method based on Kalman filter identification, which is used for suppressing positioning errors in a multi-INS navigation system equipped with more than three INS.
[0030] An analysis is conducted on a multi-INS navigation system configured with more than three INS. When configuring multiple INS, the general INS / satellite navigation combined navigation method can be used as a reference. A certain INS is used as a reference, and the position difference and velocity difference output by other INS are used as observation quantities to perform Kalman filtering estimation on the error parameters such as the gyro drift and accelerometer zero bias of the reference INS. The error parameter estimation effect also reflects the differences in the positioning capabilities of other INS to some extent. Therefore, the differences in other INS positioning characteristics, especially the longitude divergence characteristics, can be reversely identified by analyzing the estimation results of the multi-INS Kalman filter, laying the foundation for suppressing the longitude divergence error, thereby achieving the purpose of improving the positioning accuracy of the multi-INS configuration. Therefore, the present invention specifically includes the following steps:
[0031] Step 1: sort more than three INS and name them INS 1, 2, ... N in sequence, and output time synchronization information to each INS through the same time reference device. The time synchronization information includes TOD and second pulse. Each INS completes time synchronization using an external time reference and marks the timestamp information in the output longitude and latitude information;
[0032] Step 2: When the multi-inertial navigation system works normally, it outputs longitude and latitude information, east speed, north speed and vertical speed information, and heading, roll and pitch attitude information.
[0033] After the normal operation setting time, the normal operation time is generally selected as 24 hours. After the inertial navigation system has been operating normally for 24 hours, select one inertial navigation system as the reference inertial navigation system, and take the other inertial navigation systems other than the reference inertial navigation system as the reference inertial navigation system, and perform the following steps:
[0034] Step three, establish the state space model of the reference inertial navigation and the reference inertial navigation. In the state space model constructed by the embodiment of the present invention, the model observation quantity is the position difference and velocity difference between the reference inertial navigation and the reference inertial navigation; the model state quantity is the position difference, velocity difference, attitude difference between the reference inertial navigation and the reference inertial navigation, and the gyro drift and accelerometer zero bias of the reference inertial navigation. The state space model adopts the inertial navigation position, velocity, and attitude error equations, and uses Kalman filtering to obtain the gyro drift estimation value and accelerometer zero bias estimation value of the reference inertial navigation in the kth minute when each reference inertial navigation is observed.
[0035] The specific example is as follows: Taking INS 1 as the reference INS, the state space models of INS 2, 3...N and INS 1 are established respectively. The model observations are the position difference and velocity difference between INS 2, 3...N and INS 1; the model state quantities are the position difference, velocity difference, attitude difference between INS 2, 3...N and INS 1, as well as the gyro drift and accelerometer zero bias of INS 1. The model adopts the commonly used INS position, velocity and attitude error equations, and uses the classic Kalman filter to obtain the estimated value of the gyro drift of INS 1 in the kth minute when INS 2, 3...N are used as observations. and the accelerometer bias estimate and the accelerometer bias estimate are the estimated gyro drift values of the inertial navigation system 2 on the xyz axis in the north celestial east coordinate system, are the estimated gyro drift values of the inertial navigation N on the xyz axis of the north celestial east coordinate system, are the estimated values of the accelerometer zero bias of the inertial navigation system 2 on the xyz axis in the north celestial east coordinate system, are the estimated values of the accelerometer zero bias of the inertial navigation N on the xyz axis of the north celestial east coordinate system;
[0036] Step 4: Based on the gyro drift estimate and accelerometer bias estimate of the reference inertial navigation system at the kth minute obtained in step 3 for each reference inertial navigation system, calculate the longitude divergence direction parameter between the reference inertial navigation systems. The two reference inertial navigation systems are ranked p and q respectively, and the longitude divergence direction parameter between them at the kth minute is Specifically:
[0037]
[0038] in are the estimated gyro drift values of the inertial navigation system p on the x-axis, y-axis and z-axis in the north celestial east coordinate system, are the estimated accelerometer bias values of the inertial navigation system p on the x-axis, y-axis, and z-axis in the north celestial east coordinate system, are the estimated gyro drift values of the inertial navigation q on the x-axis, y-axis and z-axis in the north celestial east coordinate system, are the estimated accelerometer bias values of the inertial navigation system q on the x-axis, y-axis, and z-axis in the north celestial east coordinate system, respectively. The ⊕ operation is defined as: a and b are the objects before and after the symbol ⊕ is processed.
[0039] Step 5: If the longitude divergence direction parameters of the two reference inertial navigation systems at the kth minute are greater than or equal to 0, the longitude divergence directions of the two reference inertial navigation systems are the same, and it is necessary to further calculate the longitude divergence amplitude parameter between the two reference inertial navigation systems.
[0040] Calculate the longitude divergence amplitude parameter between the two reference inertial navigation systems:
[0041]
[0042] The ⊙ operation is defined as: a' and b' are the objects before and after the symbol ⊙ processing respectively.
[0043] Then, the longitude divergence direction parameter and the longitude divergence amplitude parameter are combined to fuse the longitudes of the two reference inertial navigation systems to obtain the fused longitude information for output.
[0044] , specifically:
[0045] like Then the longitude divergence directions of INS p and INS q are the same, and the longitude divergence amplitude of INS p is smaller than that of INS q. When multiple INS are fused, INS p longitude is selected as the fused longitude information output;
[0046] like Then the longitude divergence directions of INS p and INS q are the same, and the longitude divergence amplitude of INS p is greater than that of INS q. When multiple INS are fused, INS q longitude is selected as the fused longitude information output;
[0047] like The longitude divergence directions of INS p and INS q are different. When multiple INS are fused, the weighted average of the longitudes of the two INS is selected as the fused longitude information output.
[0048] Step 6: Switch the reference INS to the next INS, and use the other INS other than the reference INS as the reference INS. Repeat steps 3 to 5 until the fused longitude information between all INS in the multi-INS navigation system is determined. Select two groups of INS with different longitude divergence directions and the smallest divergence amplitude in the same direction, and take the weighted average of their output longitudes as the multi-INS fusion longitude output.
[0049] In summary, the above are only preferred embodiments of the present invention and are not intended to limit the protection scope of the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the protection scope of the present invention.
Claims
1. A multi-INS positioning error suppression method based on Kalman filter identification, It is characterized in that It is used to suppress positioning errors in a multi-INS navigation system equipped with more than three INS, and specifically includes the following steps: Step 1: sort more than three multi-INS and name them as INS 1, 2, ... N in sequence, and output time synchronization information to each INS through the same time reference device, wherein the time synchronization information includes TOD and second pulse, and each INS completes time synchronization using an external time reference, and marks timestamp information in the output longitude and latitude information; Step 2: After the normal operation of multiple inertial navigation systems for a set period of time, select one inertial navigation system as the reference inertial navigation system and take the other inertial navigation systems as the reference inertial navigation systems. Perform the following steps: Step 3, establishing a state space model of the reference inertial navigation and the reference inertial navigation; the state space model adopts the inertial navigation position, velocity, and attitude error equations, and uses Kalman filtering to obtain the gyro drift estimation value and accelerometer zero bias estimation value of the reference inertial navigation at the kth minute when each reference inertial navigation is observed; Step 4: Based on the gyro drift estimation value and accelerometer bias estimation value of the reference inertial navigation at the kth minute under each reference inertial navigation observation condition obtained in step 3, the longitude divergence direction parameters between the reference inertial navigations are calculated; the calculation of the longitude divergence direction parameters between the reference inertial navigations is specifically as follows: The order of the two reference inertial navigations is p and q respectively, then the longitude divergence direction parameter between the two at the kth minute is Specifically: in are the estimated gyro drift values of the inertial navigation system p on the x-axis, y-axis and z-axis in the north celestial east coordinate system, are the estimated accelerometer bias values of the inertial navigation system p on the x-axis, y-axis, and z-axis in the north celestial east coordinate system, are the estimated gyro drift values of the inertial navigation q on the x-axis, y-axis and z-axis in the north celestial east coordinate system, are the estimated accelerometer bias values of the inertial navigation system q on the x-axis, y-axis, and z-axis in the north celestial east coordinate system, respectively. The ⊕ operation is defined as: a and b are the front and back objects of the symbol ⊕ processing respectively; Step 5: If the longitude divergence direction parameters of the two reference inertial navigation systems at the kth minute are greater than or equal to 0, the longitude divergence directions of the two reference inertial navigation systems are the same, and it is necessary to further calculate the longitude divergence amplitude parameters between the two reference inertial navigation systems; then, the longitude divergence direction parameters and the longitude divergence amplitude parameters are combined to fuse the longitudes of the two reference inertial navigation systems to obtain the fused longitude information for output; The longitude divergence amplitude parameters between the two reference inertial navigation systems are calculated as follows: The ⊙ operation is defined as: a' and b' are the front and back objects of the symbol ⊙ processing respectively; The longitude divergence direction parameter and the longitude divergence amplitude parameter are combined to fuse the longitudes of the two reference inertial navigation systems to obtain fused longitude information for output, specifically: like Then the longitude divergence directions of INS p and INS q are the same, and the longitude divergence amplitude of INS p is smaller than that of INS q. When multiple INS are fused, INS p longitude is selected as the fused longitude information output; like Then the longitude divergence directions of INS p and INS q are the same, and the longitude divergence amplitude of INS p is greater than that of INS q. When multiple INS are fused, INS q longitude is selected as the fused longitude information output; like The longitude divergence directions of INS p and INS q are different. When multiple INS are fused, the weighted average of the longitudes of the two INS is selected as the fused longitude information output; Step 6: Switch the reference inertial navigation system to the next inertial navigation system, and use the other inertial navigation systems other than the reference inertial navigation system as the reference inertial navigation system, and repeat steps 3 to 5 until the fusion longitude information between all the inertial navigation systems in the multi-inertial navigation system is determined, and two groups of inertial navigation systems with different longitude divergence directions and the smallest divergence amplitude in the same direction are selected, and the weighted average of the longitudes output by the two groups is used as the multi-inertial navigation fusion longitude output.
2. A multi-inertial navigation positioning error suppression method based on Kalman filter identification as claimed in claim 1, It is characterized in that When the multi-inertial navigation system works normally, it outputs longitude and latitude information, east speed, north speed and vertical speed information, and heading, roll and pitch attitude information.
3. The multi-inertial navigation positioning error suppression method based on Kalman filter identification as claimed in claim 1, It is characterized in that In the state space model, the model observations are the position difference and velocity difference between the reference inertial navigation and the reference inertial navigation; the model state quantities are the position difference, velocity difference and attitude difference between the reference inertial navigation and the reference inertial navigation, and the gyro drift and accelerometer zero bias of the reference inertial navigation.
