A method for online calibration of navigation error based on redundant single-axis rotation modulation
Through the redundant single-axis rotation modulation method, the problem of errors in inertial navigation system diverge over time is solved through the redundant single-axis rotation modulation method, the problem of errors diverging over time is achieved, and the online calibration and accuracy of navigation errors are enhanced, which is enhanced by the autonomy and reliability of the system. .
Patent Information
- Application Number
- CN202311053023.4
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2023-08-18
- Publication Date
- 2024-11-05
- Estimated Expiration
- 2043-08-18
AI Technical Summary
The error of the inertial navigation system diverges over time under long flights, resulting in a decrease in navigation error accuracy. The existing single-axis rotation modulation is difficult to suppress errors. The dual-axis rotation modulation is costly and large in size, which restricts the accuracy.
Using redundant single-axis rotation modulation method, the Kalman filter is designed using the position difference, speed difference and attitude constraint relationship of different rotational modulation schemes to construct the output model and sensor of the inertial measurement unit. Error model to realize online calibration and correction of navigation errors.
It improves the accuracy and reliability of the inertial navigation system, realizes real-time correction and autonomy of navigation errors, and enhances the system's autonomous navigation capabilities.
Smart Images

Figure CN117129013B_ABST
Abstract
Description
Technical Field
[0001] The invention relates to the field of inertial navigation error correction, and in particular to an online calibration method for navigation errors based on redundant single-axis rotation modulation. Background Art
[0002] In view of the advantages of inertial navigation being fully autonomous and all-weather, the carrier mainly relies on the inertial navigation system for navigation in autonomous navigation mode. Since the error of the inertial navigation system diverges over time, the navigation error accuracy decreases under long flight time. Rotational modulation technology is usually used to achieve error suppression through error self-compensation. The single-axis rotation modulation scheme cannot achieve error suppression of the modulated axis, and it is difficult to completely suppress the error in the direction of the unmodulated axis. Although the dual-axis rotation modulation can suppress errors in all directions, it is costly and bulky, and the error still diverges over time, which restricts the improvement of the inertial system accuracy. Summary of the invention
[0003] In view of this, the present invention provides a method for online calibration of navigation errors based on redundant single-axis rotation modulation. Combined with the characteristic that the single-axis rotation modulation inertial navigation system can excite the error signal, the present invention can solve the problem of online calibration of inertial navigation errors, and obtain better navigation information according to the estimated error correction redundant navigation system. The present invention can be applied and promoted in the field of inertial navigation error correction.
[0004] A navigation error online calibration method based on redundant single-axis rotation modulation comprises the following steps:
[0005] Step S1: The two sets of single-axis rotation modulation inertial navigation systems carried by the carrier move with different rotation modulation schemes. The two sets of inertial components under different modulation schemes respectively measure and obtain raw data. An inertial measurement unit output model is constructed based on the redundant single-axis rotation modulation system. At the same time, the two sets of systems respectively rotate the modulation and demodulation algorithm to solve and obtain the inertial information of the carrier system, which specifically includes:
[0006] Step S11: Two single-axis rotation modulation inertial navigation systems are installed on the same carrier and at a determined position to form a redundant single-axis rotation modulation system. The two inertial navigation systems measure the attitude, velocity, and true position of the same carrier. Each single-axis rotation modulation inertial navigation system contains a six-degree-of-freedom inertial measurement unit and a photoelectric encoder to output raw data, which is solved in the navigation solution module.
[0007] The two inertial navigation systems are inertial navigation system A and inertial navigation system B with similar sensor performance. The two systems of the rotation modulation inertial navigation system are not divided into master and slave, and the carrier coordinate system is consistent with the b system; the coordinate system of the inertial measurement unit A is defined as s A The inertial measurement unit B coordinate system is defined as s B According to the rotation modulation scheme s A Department, sB The system a is constantly changing relative to the system b, and the inertial measurement unit output model is constructed based on the redundant single-axis rotation modulation system:
[0008]
[0009]
[0010] Construct the photoelectric encoder output model:
[0011]
[0012]
[0013] and is the true value of the three-axis gyroscope in the inertial measurement unit A and the inertial measurement unit B; and They are s A Department, s B The angular velocity of the system under the frame relative to the inertial frame; and They are s A Department, s B Download system relative to s A Department, s B The angular velocity of the system, that is, the modulation angular velocity of the rotation modulation scheme; and They are the outputs of the photoelectric encoders of the inertial navigation system A and the inertial navigation system B, that is, measuring the relative rotation angle between the inertial measurement unit and the carrier system; and are the rotation angle increments of inertial navigation system A and inertial navigation system B respectively;
[0014] Step S12: Based on the raw data, the angular increment of the carrier system and the download system relative to the inertial system is calculated using a rotation modulation and demodulation algorithm; based on the inertial measurement unit output model and the photoelectric encoder output model obtained in the previous step, the angular increment of the carrier system and the download system relative to the inertial system can be calculated from the raw data output by the inertial measurement unit and the photoelectric encoder by the following formula:
[0015]
[0016]
[0017] in, They are A Department and s B Transformation matrix from system to carrier system; and The inertial navigation A and B are the outputs of the photoelectric encoder; calculated and are the angular increments of the loading system and the downloading system relative to the inertial system A and the inertial system B respectively;
[0018] Step S13: Based on the angular increment of the carrier system and the download system relative to the inertial system and the acceleration of the carrier system, the strapdown inertial navigation solution algorithm is used to solve and obtain navigation information;
[0019] Step S2: construct sensor error models for the two systems based on the rotation modulation scheme, and construct an error transfer model using the sensor rotation modulation and demodulation algorithm and the strapdown navigation solution algorithm to describe the mathematical relationship between the raw data errors and the navigation information errors of the two systems, specifically including:
[0020] S21: The error models of the two inertial measurement unit modulation outputs are constructed as follows:
[0021]
[0022]
[0023]
[0024]
[0025] in, and are the measured values of the three-axis gyroscopes in the inertial measurement unit A and the inertial measurement unit B; and is the true value of the three-axis gyroscope in the inertial measurement unit A and the inertial measurement unit B; and It is the measured value of the three axes in the inertial measurement unit A and the inertial measurement unit B; and is the true value of the three-axis accelerometer in IMU A and IMU B;
[0026] The sensor error model is constructed as follows:
[0027]
[0028]
[0029]
[0030]
[0031] In the formula, and are the three-axis gyro errors in IMU A and IMU B, and They are the three-axis added errors in inertial measurement unit A and inertial measurement unit B respectively; is the constant error of the three-axis gyroscope in IMU A and IMU B, where are the constant errors of the three axes in the inertial measurement unit A and the inertial measurement unit B respectively; and are the noise errors of the three-axis sensors in IMU A and IMU B,
[0032] Step S22: Based on the error transfer principle of strapdown inertial navigation solution, the transfer error of the inertial measurement unit error in navigation solution is established, and the transfer error equation is described in matrix form, as follows:
[0033]
[0034]
[0035]
[0036] in, represents the state transition matrix between the carrier system and the navigation system, represents the state conversion matrix between the rotation modulation coordinate system s and the carrier system b, φ is the misalignment angle, δV n =[δV E δV N ] is the speed error, δV E is the eastward velocity error, δV N is the north velocity error; δP=[δLδλ] T is the position error, L and λ are latitude and longitude respectively; M1, M2, M3 and M4 are both process matrices of strapdown inertial navigation solution, which are specifically expressed as follows:
[0037]
[0038]
[0039] Where R is the radius of the earth; ω ie is the Earth's rotation angular velocity; V E V is the carrier's geographic eastward velocity; N is the carrier's geographic northward velocity; is the specific force in the geographic coordinate system, f E is the eastward relative force, f U For the celestial force, f N For the north direction;
[0040] Step S23: The rotating mechanism controls the inertial measurement unit A to rotate around s A The x-axis direction of the system is modulated at a set angular rate, and the rotation mechanism controls the inertial measurement unit B to rotate around s B The z-axis direction of the system is modulated at a set angular rate;
[0041] The sensor error transfer model under the load system is:
[0042]
[0043]
[0044] in, is the inertial measurement unit A at s A The error propagation model of the system, is the inertial measurement unit B at s B The error transmission model of the system; ω rA and ω rB are the rotation angular velocities of the single-axis rotation mechanisms of the inertial measurement unit A and the inertial measurement unit B respectively;
[0045] Step S3: using the Kalman filter to estimate the latitude and longitude errors corresponding to the redundant inertial navigation system, establishing an error joint state equation based on the error transfer model, using the misalignment angle difference, velocity difference, and position difference between the redundant inertial navigation systems as observation quantities to establish a measurement equation, and then establishing a Kalman filter of an error online calibration model, the error online calibration model is used to online estimate the navigation parameter errors of the redundant inertial navigation system, including:
[0046] Step S31: Using the three misalignment angles of the inertial measurement unit A East and North velocity error Longitude and latitude error Three gyroscopes always drift Two plus constant drift Three misalignment angles of IMU B East and North velocity error Longitude and latitude error Three gyroscopes always drift Two plus constant drift A total of 24 state quantities are used as system state quantities. The state equation is the rotation modulation navigation error. The state equation is obtained from formulas (15)-(17) as follows:
[0047] X=[X A X B ]
[0048]
[0049]
[0050]
[0051] F is the state transfer matrix of the Kalman filter with a size of 24*24, G is the system structure parameter matrix with known parameters with a size of 24*10, and W is the system noise vector with a size of 10*1;
[0052]
[0053]
[0054]
[0055] In the above formula, is the angular velocity of the navigation system relative to the inertial system in the navigation system; The state conversion matrix from the carrier system to the navigation coordinate system n solved by the inertial system A; Representation Matrix The element at row a and column b; Represents the state conversion matrix from the carrier system to the navigation coordinate system n solved by the inertial system B; Representation Matrix The element at row a and column b;
[0056]
[0057] Step S32: Taking the position longitude and latitude of the two strapdown inertial navigation systems as the difference, and the attitude as the constraint condition, there are 7 observation quantities in total, so there are measurement vectors: the attitude constraint is calculated by multiplying the state conversion matrices calculated by the inertial navigation system A and the inertial navigation system B respectively, and the east velocity difference V EA -V EB , North velocity difference V NA -V NB , the latitude difference L1-L2, the longitude difference λ1-λ2, the measurement equation is as follows:
[0058] Z(k)=H·X(k)+V(k) (26)
[0059] Z=[L1-L2 λ1-λ2 V EA -V EB V NA -V NB O] T (27)
[0060]
[0061] O = [C(3,2) C(1,3) C(2,1)];
[0062] Among them, X(k) is the state matrix at time k, V(k) is the measurement noise matrix of size 7*1; C(a,b) represents the element at row a and column b in matrix C; φ A ,φ B are the misalignment angles of inertial navigation system A and inertial navigation system B respectively; φ A (a), φ B (a) represents the vector φ A ,φ B The ath element in ; According to formula (28), the relationship between the matrix C and the misalignment angle of the two systems is obtained, and the three elements in the matrix C are selected to form the matrix O as the constraint condition; H is the measurement equation parameter matrix of size 7*24:
[0063]
[0064] The velocity error and position error of the two systems and the constant error of the inertial measurement unit are estimated by Kalman filtering, and the navigation calculation results of the two systems are compensated in real time, thereby realizing online calibration of the navigation error.
[0065] Preferably, in step S23, the rotating mechanism controls the inertial measurement unit A to rotate around s A The x-axis direction of the system is modulated at a fixed modulation angular rate of -10° / s.
[0066] Preferably, in step S23, the rotating mechanism controls the inertial measurement unit B to rotate around s B The z-axis direction of the system is modulated at a fixed modulation angular rate of -10° / s.
[0067] Compared with the prior art, the technical solution of the present invention has the following beneficial technical effects:
[0068] 1. The present invention utilizes a redundant single-axis rotation modulation inertial navigation system to maintain autonomy while improving system reliability. By constructing a redundant rotation modulation scheme, the position difference, speed difference, and attitude constraint relationship between inertial navigation systems running with different rotation modulation schemes are used as observation quantities to design a Kalman filter, thereby realizing the longitude and latitude position error estimation of the redundant single-axis rotation modulation inertial navigation system, effectively correcting the inertial navigation information, and improving the accuracy of the inertial navigation system.
[0069] 2. The present invention can realize the real-time and complete autonomy of the inertial navigation system error estimation. BRIEF DESCRIPTION OF THE DRAWINGS
[0070] Figure 1 The figure is a flow chart of a method for online calibration of navigation error for redundant single-axis rotation modulation according to an embodiment of the present invention.
[0071] Figure 2Schematic diagram of the architecture of an online calibration method for navigation error for redundant single-axis rotation modulation according to an embodiment of the present invention.
[0072] Figure 3 FIG. 1 is a schematic diagram of a mechanical arrangement for a redundant single-axis rotation modulation scheme according to an embodiment of the present invention. DETAILED DESCRIPTION
[0073] The technical solution of the present invention is further described below in conjunction with the accompanying drawings and embodiments.
[0074] The present invention proposes an online calibration method for navigation errors of redundant single-axis rotation modulation. The present invention uses the principle that redundant inertial navigation system errors can be stimulated under different rotation modulation schemes, configures two rotation modulation inertial navigation systems, constructs a sensor error transfer model under the rotation modulation inertial navigation solution algorithm, and uses the position difference, speed difference, and attitude constraint between the inertial navigation systems under different rotation modulation schemes as observation quantities to design a Kalman filter. By estimating and correcting various errors of the inertial navigation system, the redundant inertial navigation system is used to calibrate the navigation error online, effectively solving the problem of improving the accuracy of the inertial navigation system. Figure 1-Figure 2 As shown, the online calibration method of navigation error of redundant single-axis rotation modulation of the present invention comprises the following steps:
[0075] Step S1: The two sets of single-axis rotation modulation inertial navigation systems carried by the carrier move with different rotation modulation schemes, and the two sets of inertial components under different modulation schemes are respectively measured to obtain raw data, and an inertial measurement unit output model is constructed based on the redundant single-axis rotation modulation system. At the same time, the two sets of systems are respectively rotated and modulated by demodulation algorithms to obtain the inertial information of the carrier system. Based on the inertial information of the carrier system, the two sets of navigation information of the carrier platform can be obtained by using the strapdown navigation solution algorithm, thereby forming a redundant inertial navigation system, which specifically includes:
[0076] Step S11: Two single-axis rotation modulation inertial navigation systems are installed on the same carrier and the position is determined to form a redundant single-axis rotation modulation system. The two inertial navigation systems measure the attitude, velocity, and true position of the same carrier. Each single-axis rotation modulation inertial navigation system contains a six-degree-of-freedom inertial measurement unit and a photoelectric encoder to output raw data, which is solved in the navigation solution module.
[0077] The inertial sensors used in the two inertial navigation systems have similar performances, namely inertial navigation system A and inertial navigation system B. The two systems of the rotary modulation inertial navigation system are not divided into master and slave. The carrier coordinate system is consistent with the b system, and the coordinate system is defined by the right front upper. The coordinate system of the inertial measurement unit A is defined as s A The inertial measurement unit B coordinate system is defined as s B According to the rotation modulation scheme s A Department, s BThe a-axis coordinate system is constantly transformed relative to the b-axis coordinate system, and the two coordinate systems change in a certain pattern. The output model of the inertial measurement unit is constructed based on the redundant single-axis rotation modulation system:
[0078]
[0079]
[0080] Construct the photoelectric encoder output model:
[0081]
[0082]
[0083] and is the true value of the three-axis gyroscope in the inertial measurement unit A and the inertial measurement unit B; and They are s A Department, s B The angular velocity of the system under the frame relative to the inertial frame; and They are s A Department, s B Download system relative to s A Department, s B The angular velocity of the system is also the modulation angular velocity of the rotation modulation scheme; and They are the outputs of the photoelectric encoders of the inertial navigation system A and the inertial navigation system B, respectively, to measure the relative rotation angle between the inertial measurement unit and the carrier system; and are the rotation angle increments of inertial navigation system A and inertial navigation system B respectively.
[0084] Step S12: Based on the raw data, the rotation modulation and demodulation algorithm is used to calculate the angular increment of the carrier system and the download system relative to the inertial system. Based on the inertial measurement unit output model and the photoelectric encoder output model obtained in the previous step, the raw data output by the inertial measurement unit and the photoelectric encoder can be used to calculate the angular increment of the carrier system and the download system relative to the inertial system by the following formula.
[0085]
[0086]
[0087] They are A Department and s B Transformation matrix from system to carrier system; and The inertial navigation A and B are the outputs of the photoelectric encoder; calculated and are the angular increments of the loading system and the downloading system relative to inertial system A and inertial system B respectively.
[0088] Step S13: Based on the angular increment of the carrier system and the download system relative to the inertial system and the carrier system acceleration, a strapdown inertial navigation solution algorithm is used to solve and obtain navigation information.
[0089] Step S2: construct sensor error models for the two systems based on the rotation modulation scheme, and construct an error transfer model using the sensor rotation modulation and demodulation algorithm and the strapdown navigation solution algorithm to describe the mathematical relationship between the raw data errors and the navigation information errors of the two systems, specifically including:
[0090] S21: The error models of the two inertial measurement unit modulation outputs are constructed as follows:
[0091]
[0092]
[0093]
[0094]
[0095] and are the measured values of the three-axis gyroscopes in the inertial measurement unit A and the inertial measurement unit B; and is the true value of the three-axis gyroscope in the inertial measurement unit A and the inertial measurement unit B; and It is the measured value of the three axes in the inertial measurement unit A and the inertial measurement unit B; and It is the true value of the three axes in IMU A and IMU B.
[0096] The sensor error model is constructed as follows:
[0097]
[0098]
[0099]
[0100]
[0101] In the formula, and are the three-axis gyro errors in IMU A and IMU B, and They are the three-axis added errors in inertial measurement unit A and inertial measurement unit B respectively; is the constant error of the three-axis gyroscope in IMU A and IMU B, where are the constant errors of the three axes in the inertial measurement unit A and the inertial measurement unit B respectively; and are the noise errors of the three-axis sensors in IMU A and IMU B,
[0102] Step S22: Based on the error transfer principle of strapdown inertial navigation solution, the transfer error of the inertial measurement unit error in navigation solution is established. For the convenience of analysis, the transfer error equation is described in matrix form, as follows:
[0103]
[0104]
[0105]
[0106] in, represents the state transition matrix between the carrier system and the navigation system, represents the state conversion matrix between the rotation modulation coordinate system (s system) and the carrier system (b system), φ is the misalignment angle, δV n =[δV E δV N ] is the speed error, δV E is the eastward velocity error, δV N is the north velocity error; δP=[δL δλ] T is the position error, L and λ are latitude and longitude respectively; M1, M2, M3 and M4 are both process matrices of strapdown inertial navigation solution, which are specifically expressed as follows:
[0107]
[0108]
[0109] Where R is the radius of the earth; ω ie is the Earth's rotation angular velocity; V E V is the carrier's geographic eastward velocity; N is the carrier's geographic northward velocity; is the specific force in the geographic coordinate system, f E is the eastward relative force, f U For the celestial force, f N For the north direction.
[0110] Step S23: The rotation modulation scheme is implemented, and the rotation mechanism controls the inertial measurement unit A to rotate around s A The x-axis direction of the system is modulated at a fixed modulation angular rate of -10° / s, and the rotation mechanism controls the inertial measurement unit B to rotate around s B The z-axis direction of the system is modulated at a fixed modulation angular rate of -10° / s.
[0111] The sensor error transfer model under the load system is:
[0112]
[0113]
[0114] is the inertial measurement unit A at s A The error propagation model of the system, is the inertial measurement unit B at s B The error transmission model of the system. It can be seen that ε under the rotation modulation inertial technology b Related to the rotation angle of the rotating mechanism. ω rA and ω rB are the rotation angular velocities of the single-axis rotation mechanisms of inertial measurement unit A and inertial measurement unit B respectively. The rotation mechanism excitation error makes the constant error observable.
[0115] Step S3: using the Kalman filter to estimate the latitude and longitude errors corresponding to the redundant inertial navigation system, establishing an error joint state equation based on the error transfer model, using the misalignment angle difference, velocity difference, and position difference between the redundant inertial navigation systems as observation quantities to establish a measurement equation, and then establishing a Kalman filter of an error online calibration model, wherein the error online calibration model is used to online estimate the navigation parameter errors of the redundant inertial navigation system, including:
[0116] Step S31: The position, velocity and misalignment angle of the two strapdown inertial navigation systems, as well as the gyroscope bias and accelerometer bias are used as system state quantities, a total of 24 state quantities: the three misalignment angles of the inertial measurement unit A East and North velocity error Longitude and latitude error Three gyroscopes always drift Two plus constant drift Three misalignment angles of IMU B East and North velocity error Longitude and latitude error Three gyroscopes always drift Two plus constant drift The state equation is the rotation modulation navigation error. The state equation is obtained from formulas (15)-(17) as follows:
[0117] X=[X A X B ]
[0118]
[0119]
[0120]
[0121] F is the state transfer matrix of the Kalman filter with a size of 24*24, G is the system structure parameter matrix with known parameters with a size of 24*10, and W is the system noise vector with a size of 10*1.
[0122]
[0123]
[0124]
[0125] In the above formula, is the angular velocity of the navigation system relative to the inertial system in the navigation system. The load system solved for the inertial system A (defined as b A The state conversion matrix from the navigation coordinate system (n system) to the navigation coordinate system (n system); Representation Matrix The element at row a and column b; represents the load system solved by the inertial system B (defined as b B The state conversion matrix from the navigation coordinate system (n system) to the navigation coordinate system (n system); Representation Matrix The element at row a and column b;
[0126]
[0127] Step S32: Taking the position longitude and latitude of the two strapdown inertial navigation systems as the difference, and the attitude as the constraint condition, there are 7 observation quantities in total, so there are measurement vectors: the attitude constraint is calculated by multiplying the state conversion matrices calculated by the A inertial navigation system and the B inertial navigation system respectively, and the east velocity difference V EA -V EB , North velocity difference V NA -V NB , latitude difference L1-L2, longitude difference λ1-λ2. The measurement equation is as follows:
[0128] Z(k)=H·X(k)+V(k) (26)
[0129] Z=[L1-L2 λ1-λ2 V EA -VEB V NA -V NB O] T (27)
[0130]
[0131] O=[C(3,2)C(1,3)C(2,1)]
[0132] Among them, X(k) is the state matrix at time k, and V(k) is the measurement noise matrix of size 7*1. Matrix C is calculated from the state transition matrices of the two systems, and C(a,b) represents the element at row a and column b in matrix C; φ A ,φ B are the misalignment angles of inertial navigation system A and inertial navigation system B respectively; φ A (a), φ B (a) represents the vector φ A ,φ B The ath element in the equation (28) can be used to obtain the relationship between the matrix C and the misalignment angle of the two systems. Three elements in the matrix C are selected to form the matrix O as the constraint condition. H is the measurement equation parameter matrix of size 7*24:
[0133]
[0134] The velocity error and position error of the two systems and the constant error of the inertial measurement unit are estimated by Kalman filtering, and the navigation calculation results of the two systems are compensated in real time to achieve the purpose of online calibration of navigation errors.
[0135] The above specific embodiments only describe the design principles of the present invention. The specific names in the description may be different and are not limited. Therefore, those skilled in the art in the field of the present invention may modify or replace the technical solutions recorded in the above embodiments; and these modifications and replacements do not deviate from the creative purpose and technical solutions of the present invention and should all fall within the protection scope of the present invention.
Claims
1. A navigation error online calibration method based on redundant single-axis rotation modulation, characterized in that: The steps include: Step S1: The two sets of single-axis rotation modulation inertial navigation systems carried by the carrier move with different rotation modulation schemes. The two sets of inertial components under different modulation schemes respectively measure and obtain raw data. An inertial measurement unit output model is constructed based on the redundant single-axis rotation modulation system. At the same time, the two sets of systems respectively rotate the modulation and demodulation algorithm to solve and obtain the inertial information of the carrier system, which specifically includes: Step S11: Two single-axis rotation modulation inertial navigation systems are installed on the same carrier and at a determined position to form a redundant single-axis rotation modulation system. The two inertial navigation systems measure the attitude, velocity, and true position of the same carrier. Each single-axis rotation modulation inertial navigation system contains a six-degree-of-freedom inertial measurement unit and a photoelectric encoder to output raw data, which is solved in the navigation solution module. The two inertial navigation systems are inertial navigation system A and inertial navigation system B with similar sensor performance. The two systems of the rotation modulation inertial navigation system are not divided into master and slave, and the carrier coordinate system is consistent with the b system; the coordinate system of the inertial measurement unit A is defined as s A The inertial measurement unit B coordinate system is defined as s B According to the rotation modulation scheme s A Department, s B The system a is constantly changing relative to the system b, and the inertial measurement unit output model is constructed based on the redundant single-axis rotation modulation system: Construct the photoelectric encoder output model: and is the true value of the three-axis gyroscope in the inertial measurement unit A and the inertial measurement unit B; and They are s A Department, s B The angular velocity of the system under the frame relative to the inertial frame; and They are s A Department, s B Download system relative to s A Department, s B The angular velocity of the system, that is, the modulation angular velocity of the rotation modulation scheme; and They are the outputs of the photoelectric encoders of the inertial navigation system A and the inertial navigation system B, that is, measuring the relative rotation angle between the inertial measurement unit and the carrier system; and are the rotation angle increments of inertial navigation system A and inertial navigation system B respectively; Step S12: Based on the raw data, the angular increment of the carrier system and the download system relative to the inertial system is calculated using a rotation modulation and demodulation algorithm; based on the inertial measurement unit output model and the photoelectric encoder output model obtained in the previous step, the angular increment of the carrier system and the download system relative to the inertial system can be calculated from the raw data output by the inertial measurement unit and the photoelectric encoder by the following formula: in, They are A Department and s B Transformation matrix from system to carrier system; and are the rotation angle increments of inertial navigation system A and inertial navigation system B respectively; calculated and are the angular increments of the loading system and the downloading system relative to the inertial system A and the inertial system B respectively; Step S13: Based on the angular increment of the carrier system and the download system relative to the inertial system and the acceleration of the carrier system, the strapdown inertial navigation solution algorithm is used to solve and obtain navigation information; Step S2: construct sensor error models for the two systems based on the rotation modulation scheme, and construct an error transfer model using the sensor rotation modulation and demodulation algorithm and the strapdown navigation solution algorithm to describe the mathematical relationship between the raw data errors and the navigation information errors of the two systems, specifically including: S21: The error models of the two inertial measurement unit modulation outputs are constructed as follows: in, and are the measured values of the three-axis gyroscopes in the inertial measurement unit A and the inertial measurement unit B; and is the true value of the three-axis gyroscope in the inertial measurement unit A and the inertial measurement unit B; and It is the measured value of the three axes in the inertial measurement unit A and the inertial measurement unit B; and is the true value of the three-axis accelerometer in IMU A and IMU B; The sensor error model is constructed as follows: In the formula, and are the three-axis gyro errors in IMU A and IMU B, and They are the three-axis added errors in inertial measurement unit A and inertial measurement unit B respectively; is the constant error of the three-axis gyroscope in IMU A and IMU B, where are the constant errors of the three axes added in the inertial measurement unit A and the inertial measurement unit B respectively; and are the noise errors of the three-axis sensors in IMU A and IMU B, Step S22: Based on the error transfer principle of strapdown inertial navigation solution, the transfer error of the inertial measurement unit error in navigation solution is established, and the transfer error equation is described in matrix form, as follows: in, represents the state transition matrix between the carrier system and the navigation system, represents the state conversion matrix between the rotation modulation coordinate system s and the carrier system b, φ is the misalignment angle, δV n =[δV E δV N ] is the speed error, δV E is the eastward velocity error, δV N is the north velocity error; δP=[δL δλ] T is the position error, L and λ are latitude and longitude respectively; M1, M2, M3 and M4 are both process matrices of strapdown inertial navigation solution, which are specifically expressed as follows: Where R is the radius of the earth; ω ie is the Earth's rotation angular velocity; V E V is the carrier's geographic eastward velocity; N is the carrier's geographic northward velocity; is the specific force in the geographic coordinate system, f E is the eastward relative force, f U For the celestial force, f N For the north direction; Step S23: The rotating mechanism controls the inertial measurement unit A to rotate around s A The x-axis direction of the system is modulated at a set angular rate, and the rotation mechanism controls the inertial measurement unit B to rotate around s B The z-axis direction of the system is modulated at a set angular rate; The sensor error transfer model under the load system is: in, is the inertial measurement unit A at s A The error propagation model of the system, is the inertial measurement unit B at s B The error transmission model of the system; ω rA and ω rB are the rotation angular velocities of the single-axis rotation mechanisms of the inertial measurement unit A and the inertial measurement unit B respectively; Step S3: using a Kalman filter to estimate the latitude and longitude errors corresponding to the redundant inertial navigation system, establishing an error joint state equation based on the error transfer model, using the misalignment angle difference, velocity difference, and position difference between the redundant inertial navigation systems as observation quantities to establish a measurement equation, and then establishing a Kalman filter of an error online calibration model, the error online calibration model is used to online estimate the navigation parameter errors of the redundant inertial navigation system, including: Step S31: Using the three misalignment angles of the inertial measurement unit A East and North velocity error Position error Three gyroscopes always drift Two plus constant drift Three misalignment angles of IMU B East and North velocity error Position error Three gyroscopes always drift Two plus constant drift A total of 24 state quantities are used as system state quantities. The state equation is the rotation modulation navigation error. The state equation is obtained from formulas (15)-(17) as follows: F is the state transfer matrix of the Kalman filter with a size of 24*24, G is the system structure parameter matrix with known parameters with a size of 24*10, and W is the system noise vector with a size of 10*1; In the above formula, is the angular velocity of the navigation system relative to the inertial system in the navigation system; The state conversion matrix from the carrier system to the navigation coordinate system n solved by the inertial system A; Representation Matrix The element at row a and column b; Represents the state conversion matrix from the carrier system to the navigation coordinate system n solved by the inertial system B; Representation Matrix The element at row a and column b; Step S32: Taking the position longitude and latitude of the two strapdown inertial navigation systems as the difference, and the attitude as the constraint condition, there are 7 observation quantities in total, so there are measurement vectors: the attitude constraint is calculated by multiplying the state conversion matrices calculated by the inertial navigation system A and the inertial navigation system B respectively, and the east velocity difference V EA -V EB , North velocity difference V NA -V NB , the latitude difference L1-L2, the longitude difference λ1-λ2, the measurement equation is as follows: Z(k)=H·X(k)+V(k) (26) Z=[L1-L2 λ1-λ2 V EA -V EB V NA -V NB [O] T (27) O = [C(3,2)C(1,3)C(2,1)]; Among them, X(k) is the state matrix at time k, V(k) is the measurement noise matrix of size 7*1; C(a,b) represents the element at row a and column b in matrix C; φ A ,φ B are the misalignment angles of inertial navigation system A and inertial navigation system B respectively; φ A (a), φ B (a) represents the vector φ A ,φ B The ath element in ; According to formula (28), the relationship between the matrix C and the misalignment angle of the two systems is obtained, and the three elements in the matrix C are selected to form the matrix O as the constraint condition; H is the measurement equation parameter matrix of size 7*24: The velocity error and position error of the two systems and the constant error of the inertial measurement unit are estimated by Kalman filtering, and the navigation calculation results of the two systems are compensated in real time, thereby realizing online calibration of the navigation error.
2. The method for online calibration of navigation error based on redundant single-axis rotation modulation according to claim 1, characterized in that: In step S23, the rotating mechanism controls the inertial measurement unit A to rotate around s A The x-axis direction of the system is modulated at a fixed modulation angular rate of -10° / s.
3. The method for online calibration of navigation error based on redundant single-axis rotation modulation according to claim 2, characterized in that: In step S23, the rotating mechanism controls the inertial measurement unit B to rotate around B The z-axis direction of the system is modulated at a fixed modulation angular rate of -10° / s.
