High-precision inertial navigation system static base initial alignment method based on double inertial measurement units

Abstract

The invention relates to an alignment method, in particular to a high-precision inertial navigation system static base initial alignment method based on double inertial measurement units, and solves the problems that when an existing inertial navigation system is subjected to initial alignment under the static base condition, the precision is reduced, the stability is insufficient and the cost is relatively high under the environments of vibration, temperature drift and the like. Before formal alignment, the X-axis and the Y-axis of the first inertial measurement unit and the second inertial measurement unit are fixedly mounted after being reversed, so that error terms of the first inertial measurement unit and the second inertial measurement unit have the characteristic of opposite symbols in the horizontal direction, a physical implementation basis is provided for common-mode error suppression, and the influence caused by mounting errors is reduced through multiple independent alignment; the alignment precision is improved; during formal alignment, the coarse alignment results of the two inertial measurement units are fused through the relative attitude matrix, the two inertial measurement units respectively use the fused coarse alignment results to carry out fine alignment, and exchange alignment is carried out again in the fine alignment process.

Description

TECHNICAL FIELD

[0001] The application relates to an alignment method, in particular to a high-precision inertial navigation system static base initial alignment method based on a double inertial measurement unit. BACKGROUND

[0002] Currently, the initial alignment of a strapdown inertial navigation system is a key process for determining the initial attitude between a carrier and a navigation coordinate system, and the precision and stability thereof directly determine the precision of the inertial navigation result.

[0003] The initial alignment is divided into static base alignment and dynamic alignment, wherein the static base alignment refers to the alignment of a carrier in a static state only by relying on the earth rotation angular velocity and the earth gravity field.

[0004] The static base alignment is generally divided into two stages of coarse alignment and fine alignment, the coarse alignment process quickly estimates the attitude by analyzing the gravity vector and the earth rotation vector, and the precision thereof is limited; and the fine alignment process generally adopts a Kalman filter algorithm to further optimize the attitude and estimate the device error, so as to realize the acquisition of the carrier attitude with high precision.

[0005] The current commonly used schemes for the static base alignment of the strapdown inertial navigation system are generally a single inertial measurement unit scheme and a mechanical turntable auxiliary scheme.

[0006] 1) Single inertial measurement unit scheme: Generally, an analytical coarse alignment and a Kalman filter fine alignment or a least square fine alignment are adopted. In the coarse alignment, the initial attitude matrix is directly calculated by using the gravity vector and the earth rotation vector; and in the fine alignment process, the coarse alignment result is taken as an initial value, and the attitude error angle and the device zero offset are estimated by using the velocity and position error observation.

[0007] This scheme has simple structure and small calculation amount, but due to the precision limitation of the inertial measurement device and the small magnitude of the earth rotation angular velocity, the heading angle measurement error is large, and the precision will be significantly reduced under the environment of vibration, temperature drift and the like.

[0008] 2) Mechanical turntable auxiliary scheme Generally, a turntable rotation modulation or a multi-position data fusion is used. In the turntable rotation modulation, the initial attitude matrix is still directly calculated by using the gravity vector and the earth rotation vector in the coarse alignment; but in the fine alignment process, the inertial measurement unit is rotated by the turntable, the inertial device error is modulated as a sine function, and the error is eliminated by integration. In the multi-position data fusion, the inertial measurement unit is rotated by 180 degrees in the fine alignment process, so as to improve the observability of the system and the alignment precision.

[0009] This scheme has high alignment accuracy, but due to the high-precision turntable is expensive and difficult to maintain, will significantly increase the cost, and the failure rate is higher. SUMMARY

[0010] In order to solve the problem that the existing inertial navigation system has low precision and poor stability under the condition of vibration, temperature drift and other environments when performing initial alignment under static base, and the cost is high, the application provides a high-precision inertial navigation system static base initial alignment method based on double inertial measurement units.

[0011] In order to achieve the above purpose, the application adopts the following technical scheme: A high-precision inertial navigation system static base initial alignment method based on double inertial measurement units, characterized in that: S1, two inertial measurement units are named as first inertial measurement unit and second inertial measurement unit, and the X axis and Y axis of the two are fixedly installed after being reversed, so that the positive direction of the X axis of the first inertial measurement unit is opposite to that of the X axis of the second inertial measurement unit, the positive direction of the Y axis of the first inertial measurement unit is opposite to that of the Y axis of the second inertial measurement unit, and the Z axes of the first inertial measurement unit and the second inertial measurement unit are both directed to the sky; S2, under the condition of static base, control the first inertial measurement unit and the second inertial measurement unit to perform N times of independent initial alignment respectively, each alignment includes coarse alignment and fine alignment, and output the attitude matrix obtained by each alignment After alignment, all attitude matrices of the first inertial measurement unit and the second inertial measurement unit are respectively The average attitude matrix is obtained by averaging ; S3, the relative attitude matrix between the first inertial measurement unit and the second inertial measurement unit is obtained through the average attitude matrix of the first inertial measurement unit and the second inertial measurement unit : ; S4, the first inertial measurement unit and the second inertial measurement unit are respectively based on the gravity vector And the earth rotation angular velocity vector Perform coarse alignment analysis, and output attitude matrix And ; S5, the attitude matrix of the first inertial measurement unit Is converted to the coordinate system of the second inertial measurement unit, and the attitude matrix of the second inertial measurement unit Is converted to the coordinate system of the first inertial measurement unit, and the conversion formula is as follows: ; ); S6, the attitude matrix of the first inertial measurement unit is calculated according to the angular velocity and the specific force of the first inertial measurement unit and the attitude matrix of the second inertial measurement unit is calculated according to the angular velocity and the specific force of the second inertial measurement unit a weighted average is performed according to the device noise variance of the first inertial measurement unit and the second inertial measurement unit to obtain a final coarse alignment result : ; ; wherein: the device noise variance of the first inertial measurement unit, the device noise variance of the second inertial measurement unit; S7, a Kalman filter is performed on the Kalman filter with the initial value of the final coarse alignment result to perform Kalman filter fine alignment, and the state vector of the Kalman filter is ; wherein: the attitude angle error, the velocity error, the gyro zero offset, the accelerometer zero offset, and T is the transpose of the matrix; The state matrix equation and the observation matrix equation of the Kalman filter are as follows: ; wherein: the state vector is the derivative with respect to time, the state transition matrix; the state vector, the noise input matrix, the process noise vector, the system observation matrix, the observation noise vector, the eastward velocity, the northward velocity, the system observation vector; S8, the angular velocity and the specific force of the first inertial measurement unit and the second inertial measurement unit are input into the Kalman filter to perform position and velocity calculation, and the velocity and position information of the first inertial measurement unit and the second inertial measurement unit are obtained respectively; ; S9, the velocity and position information of the first inertial measurement unit and the second inertial measurement unit are used to augment the observation matrix simultaneously: ; ; wherein: the augmented system observation vector,​ is an observation vector of the first inertial measurement unit, is an observation vector of the second inertial measurement unit, is an augmented system observation matrix, is an observation matrix of the first inertial measurement unit, is an observation matrix of the second inertial measurement unit; S10, obtaining the Kalman filter refined alignment results of the first inertial measurement unit and the second inertial measurement unit by the Kalman filter completed by the augmented observation matrix ; S11, for realizing equivalent double-position observation, at the midpoint of the refined alignment time, exchanging the Kalman filter refined alignment results of the first inertial measurement unit and the second inertial measurement unit at the current time: ; ; The double-inertial measurement unit-based high-precision inertial navigation system static base initial alignment is completed.

[0012] Further, in step S11, according to the attitude angle error , the and are compensated according to the following formula: .

[0013] Further, in step S2, under the static base condition, the first inertial measurement unit and the second inertial measurement unit are respectively controlled to perform more than 10 times of independent initial alignment.

[0014] Advantages of the present application: 1. The double-inertial measurement unit-based high-precision inertial navigation system static base initial alignment method provided by the present application improves the observability of the system state quantity, and effectively improves the initial alignment precision and stability compared with the inertial navigation system composed of a single inertial measurement unit; compared with the inertial navigation system assisted by a mechanical turntable, the mechanical structure cost is reduced, and there is no mechanical structure motion noise.

[0015] 2. The double-inertial measurement unit-based high-precision inertial navigation system static base initial alignment method provided by the present application reverses and then fixedly installs the X-axis and the Y-axis of the first inertial measurement unit and the second inertial measurement unit, so that the positive direction of the X-axis of the first inertial measurement unit is opposite to the positive direction of the X-axis of the second inertial measurement unit, the positive direction of the Y-axis of the first inertial measurement unit is opposite to the positive direction of the Y-axis of the second inertial measurement unit, and the Z-axes of the first inertial measurement unit and the second inertial measurement unit are both oriented to the sky, so that the error terms of the two in the horizontal direction present the characteristic of opposite signs, thereby providing a physical implementation basis for common-mode error suppression.​

[0016] 3. The high-precision inertial navigation system static base initial alignment method based on double inertial measurement units provided by the application can effectively reduce the influence of installation errors by obtaining the relative attitude matrix between the two inertial measurement units through multiple independent alignment calibration before formal alignment.

[0017] 4. The high-precision inertial navigation system static base initial alignment method based on double inertial measurement units provided by the application can significantly improve the initial alignment accuracy and stability by using rotation matrix conversion and weighted average to fuse the coarse alignment results of the two inertial measurement units and obtain the final coarse alignment result.

[0018] 5. In the Kalman filter fine alignment process of the high-precision inertial navigation system static base initial alignment method based on double inertial measurement units provided by the application, the angular velocity and specific force of the first inertial measurement unit and the second inertial measurement unit are used to solve the position and velocity, and the velocity and position information of the first inertial measurement unit and the second inertial measurement unit are obtained to augment the observation matrix, which is equivalent to realizing the double position observation effect of the mechanical turntable. BRIEF DESCRIPTION OF DRAWINGS

[0019] Figure 1 Figure 1 is a schematic diagram of the coordinate system of the first inertial measurement unit and the second inertial measurement unit after installation in the high-precision inertial navigation system static base initial alignment method based on double inertial measurement units provided by the application. DETAILED DESCRIPTION

[0020] The technical solutions of the application will be described clearly and completely below in combination with the drawings and embodiments. Obviously, the described embodiments are only part of the embodiments of the application, rather than all the embodiments. Based on the embodiments of the application, all other embodiments obtained by those skilled in the art without creative labor fall within the scope of protection of the application.

[0021] The high-precision inertial navigation system static base initial alignment method based on double inertial measurement units provided by the embodiment comprises the following steps: S1, two inertial measurement units are named as the first inertial measurement unit and the second inertial measurement unit, and the X axis and the Y axis of the two are fixedly installed after being reversed, so that the positive direction of the X axis of the first inertial measurement unit is opposite to the positive direction of the X axis of the second inertial measurement unit, the positive direction of the Y axis of the first inertial measurement unit is opposite to the positive direction of the Y axis of the second inertial measurement unit, and the Z axes of the first inertial measurement unit and the second inertial measurement unit are both directed to the sky; The coordinate system of the first inertial measurement unit and the second inertial measurement unit after installation is shown in Figure 1 ; and S2, control the first and second inertial measurement units to perform N times of independent initial alignment respectively under the static base condition, each alignment including coarse alignment and fine alignment, and output the attitude matrix obtained in each alignment After the alignment is completed, average all the attitude matrices of the first and second inertial measurement units to obtain an average attitude matrix ; In the embodiment, more than 10 times of independent initial alignment is performed to ensure the sample amount when the average is taken; S3, obtain the relative attitude matrix between the first and second inertial measurement units through the average attitude matrix of the first and second inertial measurement units : ; S4, perform analytical coarse alignment on the first and second inertial measurement units respectively based on the gravity vector and the earth rotation angular velocity vector , and output the attitude matrix and ; S5, convert the attitude matrix of the first inertial measurement unit to the coordinate system of the second inertial measurement unit, and convert the attitude matrix of the second inertial measurement unit to the coordinate system of the first inertial measurement unit, and the conversion formula is as follows: ; ); S6, perform weighted average on the attitude matrix of the first inertial measurement unit and the attitude matrix of the second inertial measurement unit according to the device noise variance of the first and second inertial measurement units to obtain the final coarse alignment result : ; ; wherein: is the device noise variance of the first inertial measurement unit, is the device noise variance of the second inertial measurement unit; S7, construct a Kalman filter with the final coarse alignment result as the initial value to perform Kalman filtering fine alignment, and the state vector ; Because in the field of inertial navigation, the Kalman filter needs to input an "initial value X0", in the application, the final coarse alignment result is taken as the "initial value X0". wherein: is the attitude angle error, is the velocity error, is the gyro bias, is the accelerometer bias, T is the transpose of a matrix; The state matrix equation and the observation matrix equation of the Kalman filter are as follows: ; wherein: is the state vector is the derivative with respect to time, is the state transition matrix; is the state vector, is the noise input matrix, is the process noise vector, is the system observation matrix; is the observation noise vector, is the eastward velocity, is the northward velocity, is the system observation vector; S8, input the angular velocity of the first inertial measurement unit and the second inertial measurement unit and the specific force to the Kalman filter, and the velocity and position information of the first inertial measurement unit and the second inertial measurement unit are obtained respectively through position and velocity calculation; S9, the velocity and position information obtained through the calculation of the first inertial measurement unit and the second inertial measurement unit are used to augment the observation matrix: ; ; wherein: is the augmented system observation vector, is the observation vector of the first inertial measurement unit, is the observation vector of the second inertial measurement unit, is the augmented system observation matrix, is the observation matrix of the first inertial measurement unit, is the observation matrix of the second inertial measurement unit; S10, the Kalman filter result of the first inertial measurement unit and the second inertial measurement unit is obtained through the Kalman filter after the completion of the augmented observation matrix ; S11, in order to realize equivalent double-position observation, at the midpoint of the fine alignment time, the Kalman filter fine alignment result of the first inertial measurement unit and the second inertial measurement unit at the current time is exchanged: ;​ ; At the same time according to the attitude angle error According to the following formula to And Compensation is carried out: ; Complete high-precision inertial navigation system static base initial alignment based on double inertial measurement unit.

[0022] The application is designed to improve the precision and stability of the strapdown inertial navigation system in the initial alignment stage, and a high-precision stable initial alignment is realized by compensating the configuration of the double non-rotating table inertial measurement unit. Before formal alignment, the X axis and Y axis of the first inertial measurement unit and the second inertial measurement unit are fixedly installed after being reversed, so that the error terms of the two in the horizontal direction present the characteristics of opposite signs, providing a physical implementation basis for common-mode error suppression. And through multiple independent alignment, the influence of installation error is reduced, and the alignment precision is improved; in formal alignment, the coarse alignment results of the two inertial measurement units are fused through the relative attitude matrix, the two inertial measurement units use the fused coarse alignment results for fine alignment respectively, and exchange alignment is carried out again in the fine alignment process, and finally the alignment is completed.

[0023] The above is only a specific embodiment of the application, but the protection scope of the application is not limited thereto, any change or replacement within the technical scope disclosed by the application should be covered within the protection scope of the application. Therefore, the protection scope of the application should be subject to the protection scope of the claims.

Claims

1. A method for initial alignment of a static base of a high-precision inertial navigation system based on dual inertial measurement units, characterized in that, Includes the following steps: S1. Name the two inertial measurement units as the first inertial measurement unit and the second inertial measurement unit, respectively, and fix them with their X-axis and Y-axis reversed, so that the positive directions of the X-axis of the first inertial measurement unit and the X-axis of the second inertial measurement unit are opposite, the positive directions of the Y-axis of the first inertial measurement unit and the Y-axis of the second inertial measurement unit are opposite, and the Z-axis of the first inertial measurement unit and the second inertial measurement unit both face upwards. S2. Under static base conditions, control the first inertial measurement unit and the second inertial measurement unit to perform N independent initial alignments respectively. Each alignment includes coarse alignment and fine alignment, and outputs the attitude matrix obtained from each alignment. After alignment, the attitude matrices of the first and second inertial measurement units are respectively... The average attitude matrix is ​​obtained by averaging. ; S3. Obtain the relative attitude matrix between the first and second inertial measurement units by averaging their attitude matrices. : ； S4. The first inertial measurement unit and the second inertial measurement unit are respectively based on the gravity vector. With the Earth's rotational angular velocity vector Perform analytical coarse alignment and output the attitude matrix respectively. and ; S5. The attitude matrix of the first inertial measurement unit. Transform to the coordinate system of the second inertial measurement unit (IMU), and obtain the attitude matrix of the second IMU. The transformation to the first inertial measurement unit coordinate system is as follows: ； ）； S6. The attitude matrix of the first inertial measurement unit. Attitude matrix with the second inertial measurement unit The final coarse alignment result is obtained by weighted averaging the device noise variances of the first and second inertial measurement units. : ； ； in: The device noise variance of the first inertial measurement unit. The device noise variance of the second inertial measurement unit; S7. Construct the final coarse alignment result. For a Kalman filter with initial values, perform Kalman filtering fine alignment, and its state vector... ; in: For attitude angle error, For speed error, For zero bias of the gyroscope, For accelerometer zero bias, T is the transpose of the matrix; The state matrix equation and observation matrix equation of the Kalman filter are as follows: ； in: State vector The derivative with respect to time, This is the state transition matrix; For state vectors, The noise input matrix, This is the process noise vector. This is the system observation matrix; To observe the noise vector, For eastward speed, For northbound speed, This represents the system observation vector; S8. Input the angular velocities of the first inertial measurement unit and the second inertial measurement unit. Compared to The Kalman filter is used to calculate the position and velocity, thereby obtaining the velocity and position information of the first inertial measurement unit and the second inertial measurement unit, respectively. S9. Simultaneously augment the observation matrix using the velocity and position information obtained from the first and second inertial measurement units: ； ； in: For the augmented system observation vector, This is the observation vector of the first inertial measurement unit. This is the observation vector of the second inertial measurement unit. The augmented system observation matrix, This is the observation matrix of the first inertial measurement unit. This is the observation matrix for the second inertial measurement unit; S10. The Kalman filter used after augmenting the observation matrix is ​​applied yields the Kalman filter-based fine alignment results between the first and second inertial measurement units. ; S11. To achieve equivalent dual-position observation, at the midpoint of the fine alignment time, analyze the Kalman-filtered fine alignment results of the first and second inertial measurement units at the current moment. To interchange: ； ； Complete the initial alignment of the static base of the high-precision inertial navigation system based on dual inertial measurement units.

2. The initial alignment method for a static base of a high-precision inertial navigation system based on dual inertial measurement units according to claim 1, characterized in that: In step S11, based on the attitude angle error According to the following formula and Compensation will be provided. 。 3. The initial alignment method for a high-precision inertial navigation system static base based on dual inertial measurement units according to claim 2, characterized in that: In step S2, under static base conditions, the first inertial measurement unit and the second inertial measurement unit are controlled to perform more than 10 independent initial alignments respectively.

Images