Combined navigation device and method based on multi-baseline redundancy observation

Abstract

The invention discloses an integrated navigation device and method based on multi-baseline redundancy observation, and belongs to the technical field of satellite navigation and inertial navigation. The method comprises the following steps: configuring a multi-antenna GNSS receiver array, acquiring high-precision space vector information by utilizing a carrier phase difference technology, realizing non-collinear multi-baseline measurement, converting an error between corresponding baselines by utilizing a plurality of non-collinear GNSS measurement baselines and an inertial navigation attitude, taking the error as an observed quantity, and constructing a combined filtering model. And rapid attitude calibration of the medium and low precision inertial navigation system is realized through Kalman filtering. According to the method, the full-dimensional attitude observability is enhanced through geometric constraints of multiple baselines, the system reliability is still kept under the condition that a single baseline fails, the attitude stability and convergence speed of a low-cost navigation system are remarkably improved, and the method can be widely applied to unmanned aerial vehicles and automatic driving.

Description

Technical Field

[0001] This invention relates to the field of satellite navigation and inertial navigation technology, specifically to a combined navigation device and method based on multi-baseline redundant observation, which is particularly suitable for high-precision attitude rapid estimation and real-time calibration of low-cost inertial navigation systems. Background Technology

[0002] Low-cost fiber optic inertial navigation systems (FOG INS) and microelectromechanical systems (MEMS INS) suffer from large random errors and system biases due to their lower sensor accuracy. This causes attitude errors to accumulate and diverge rapidly over time, making it difficult to meet the requirements for long-term, high-precision navigation. To suppress the accumulation of errors in inertial navigation systems, GNSS / INS integrated navigation technology has been widely adopted.

[0003] Traditional GNSS / INS integrated navigation schemes often employ filtering methods based on GNSS position and velocity information, or single-baseline heading angle-assisted filtering, to correct attitude errors in the inertial system. However, these methods have certain limitations: filtering strategies based on GNSS position and velocity observations cannot effectively estimate attitude and inertial device errors under static or uniform motion conditions, leading to attitude error divergence; simultaneously, in dynamic environments, the initial alignment time is long, affecting real-time response performance; while single-baseline heading angle-assisted filtering schemes cannot provide complete horizontal angle observation information, resulting in slow attitude error convergence; furthermore, single-baseline heading angle observations have low redundancy, making the system sensitive to external disturbances such as baseline loss of lock and multipath effects, resulting in insufficient overall robustness. Therefore, this paper proposes an integrated navigation method based on multi-baseline redundant observations. This method, tightly coupled with INS to construct a collaborative observation model, can effectively solve the problems of full-dimensional observability of attitude errors and rapid error estimation. Summary of the Invention

[0004] To address the aforementioned issues, this invention provides a combined navigation device and method based on multi-baseline redundant observations. It utilizes redundant carrier antenna configuration and GNSS carrier phase differential technology to acquire high-precision spatial vector information, achieving non-collinear multi-baseline measurements. Furthermore, it uses the errors of the corresponding baselines in the navigation system calculated from the multiple non-collinear GNSS measurement baselines and the inertial navigation attitude as observations to construct a compactly combined filtering model. The multi-baseline configuration provides complete observability of the full-dimensional attitude angles. Simultaneously, the redundancy of the observation information greatly enhances the system's fault tolerance and robustness, providing reliable attitude constraints even in the event of short-term failure of some baselines. This method can effectively calibrate the attitude error drift of low-precision inertial navigation systems (INS), providing a reliable solution for improving the attitude stability of low-cost navigation systems.

[0005] To achieve the above objectives, the solution of the present invention is as follows:

[0006] A combined navigation device based on multi-baseline redundant observation includes a multi-antenna GNSS receiving module, an IMU and inertial calculation module, a data fusion processing module, and a navigation output module;

[0007] The multi-antenna GNSS receiving module consists of a receiver and at least three GNSS antennas. The GNSS antennas are rigidly connected to the carrier to form multiple non-collinear short baselines, ensuring that the phase center of the GNSS antennas is anchored in the carrier coordinate system. The receiver receives satellite signals and uses carrier phase double difference observation technology and integer period ambiguity fixing method to calculate the baseline vector in the Earth coordinate system.

[0008] The IMU and inertial calculation module are used to collect the raw data of the carrier's angular velocity and specific force in real time, and to complete the inertial navigation calculation.

[0009] The data fusion processing module achieves synchronous communication with the multi-antenna GNSS receiving module, IMU, and inertial calculation module through a data interface. It constructs the Kalman filter state equation and Kalman filter observation equation to realize the real-time operation of the Kalman filtering algorithm. Based on the inertial navigation error propagation law, it constructs a 15-dimensional state equation. Based on the baseline error of the GNSS baseline measurement vector and the baseline vector calculated based on the inertial navigation attitude in the navigation coordinate system, it constructs the observation measurement. Using the Kalman filtering algorithm, it iteratively estimates each state error through the state prediction and observation update process.

[0010] The navigation output module calibrates the inertial calculation results using the state errors obtained from the data fusion processing module, and outputs the calibrated attitude, velocity, and position parameters of the inertial calculation to provide high-precision navigation information for the carrier.

[0011] A GNSS / INS integrated navigation method based on multi-baseline redundant observations, implemented through an integrated navigation device based on multi-baseline redundant observations, is characterized by the following steps:

[0012] Step 1: Set up coordinate systems and calibrate the baseline of the carrier system, including the Earth coordinate system, IMU coordinate system, carrier coordinate system and navigation coordinate system;

[0013] Step 2: The multi-antenna GNSS receiving module receives signals from more than 5 satellites, constructs a double-difference observation model, eliminates satellite clock errors and receiver clock errors, and calculates the baseline vectors of the carrier system in the Earth coordinate system. , , Simultaneously, the inertial measurement unit of the IMU and inertial calculation module outputs the raw data of the carrier's three-axis angular velocity and specific force in real time. After processing by the strapdown inertial navigation system, the carrier's attitude matrix is ​​calculated. By combining the velocity and position update equations, continuous INS navigation parameter outputs are generated, including latitude and longitude, altitude, velocity, and Euler attitude angles.

[0014] Step 3: Select a 15-dimensional state vector, including attitude misalignment angle error. Speed ​​error Position error Gyroscope zero bias error and the zero bias error of the table State vector Represented as:

[0015]

[0016] The continuous-time state transition equation is expressed as:

[0017]

[0018] In the formula, State variables Time derivative, The state transition matrix is ​​determined by the inertial navigation error propagation relationship. The noise distribution matrix is... This is the system noise vector;

[0019] Step 4: Combining the relationship between attitude misalignment angle error and baseline vector error, stack the m baselines to construct the observation matrix. and residual vector This leads to the formation of Kalman filter observation equations based on multi-baseline observations;

[0020] Step 5: Discretize the continuous-time state equation and the Kalman filter observation equation, input them into the Kalman filter, and estimate the attitude misalignment angle error through the prediction-update process. Using attitude misalignment angle error The attitude output of the INS is calibrated and updated in real time.

[0021] Furthermore, the origin of the Earth coordinate system Centered on Earth The axis lies in the equatorial plane and points towards the central meridian. The axis is along the direction of the Earth's rotation. , and The relationship between the three axes satisfies the right-hand rule;

[0022] IMU coordinate system origin Located in the sensitive center of the IMU, axis, shaft and The axes are respectively aligned with the horizontal axis to the right, the vertical axis forward, and the vertical axis upward, forming a right-front-upper coordinate system;

[0023] Carrier coordinate system: origin Located at the sensitive center of the carrier, axis, shaft and The axes are respectively positioned to the right along the transverse axis, forward along the longitudinal axis, and upward along the vertical axis of the carrier, forming a right-front-upper coordinate system;

[0024] Navigation coordinate system: The navigation coordinate system is the coordinate system selected when calculating navigation parameters. The Northeast-Heaven-Earth coordinate system is adopted as the navigation coordinate system.

[0025] The IMU coordinate system coincides with the carrier coordinate system.

[0026] Furthermore, the carrier system baseline calibration employs a laser tracker or industrial photogrammetry system to perform three-dimensional coordinate measurements of the phase center positions of each GNSS antenna, and further calculates the carrier coordinate system vectors of the multiple non-collinear short baselines formed between the GNSS antennas. , , And store.

[0027] Furthermore, the specific process of step 4 is as follows:

[0028] Based on the baseline vector data obtained through baseline calibration of the carrier coordinate system in step 1, and the carrier attitude matrix output by the inertial measurement unit of the IMU and inertial calculation module in step 2. Complete the baseline vector , , Transformation in the navigation coordinate system;

[0029] For the i-th system baseline vector It is processed by the attitude transition matrix. Projecting onto the navigation coordinate system yields the baseline vector derived by the inertial navigation system. :

[0030]

[0031] Define the baseline vector calculated in step 2 and the residual of the baseline vector derived from inertial navigation in the navigation coordinate system as the observation. For the i-th baseline, the baseline vector calculated by GNSS double-difference observation... Baseline vector corresponding to inertial navigation calculation Differential under navigation system That is, observation Its mathematical expression is:

[0032]

[0033] in, The transformation matrix from the Earth coordinate system to the navigation coordinate system is calculated based on the latitude and longitude information obtained in step 2.

[0034] Combining the relationship between attitude misalignment angle error and baseline vector error in, The residual error term will include , , An observation matrix is ​​constructed by stacking m baselines, including [the baselines mentioned above]. and residual vector And establish linearized observation equations :

[0035] .

[0036] The present invention has the following substantial advantages over the prior art:

[0037] (1) Multi-baseline redundant observation mechanism, which enhances the three-dimensional attitude of pitch angle, roll angle and heading angle to be fully observable through spatial geometric constraints;

[0038] (2) Establish a baseline error mapping model between the carrier system and the navigation system, and design a multi-baseline observation quantum stacking observation Kalman filter model to effectively reduce the attitude error convergence time and significantly improve the heading angle estimation accuracy, providing a brand-new technical solution for low-cost integrated navigation systems. Attached Figure Description

[0039] Figure 1 This is a schematic diagram of the installation of a combined navigation device based on multi-baseline redundant observation according to the present invention.

[0040] Figure 2 This is a functional composition diagram of a combined navigation device based on multi-baseline redundant observation according to the present invention.

[0041] Figure 3 This is a flowchart of an algorithm for a combined navigation method based on multi-baseline redundant observations according to the present invention. Detailed Implementation

[0042] like Figure 2 As shown, a GNSS / INS tightly coupled navigation device with multi-baseline redundant observation mainly includes a multi-antenna GNSS receiver module, an IMU (Inertial Measurement Unit) and inertial calculation module, a data fusion processing module, and a navigation output module. The functions of each module are detailed below:

[0043] Multi-antenna GNSS receiver module: Composed of no fewer than three non-collinear GNSS antennas and a receiver. The antennas are rigidly connected to the carrier, forming multiple non-collinear short baselines, and ensuring that the antenna phase center is anchored in the carrier coordinate system. The receiver receives satellite signals and uses carrier phase double-difference observation technology and integer period ambiguity fixing method to calculate the baseline vector in the Earth coordinate system. Typically, its short baseline measurement accuracy can reach the millimeter level.

[0044] IMU and Inertial Calculation Module: Low-cost inertial navigation equipment (FOG INS or MEMS INS) is used to collect the raw data of the carrier's angular velocity and specific force in real time and complete the inertial navigation calculation. The IMU coordinate system is installed in the same coordinate system as the carrier.

[0045] The data fusion processing module employs an embedded processor and a dedicated computing chip. It communicates synchronously with the multi-antenna GNSS receiver module, IMU, and inertial calculation module via a data interface. Its core functions include: constructing the state equation and observation equation for the Kalman filter, and performing real-time computation of the Kalman filtering algorithm. Based on the inertial navigation error propagation law, a 15-dimensional state equation is constructed. Observations are constructed based on the baseline error between the GNSS baseline measurement vector and the baseline vector calculated based on the inertial navigation attitude in the navigation coordinate system. Through a filtering algorithm, the errors of each state are iteratively estimated through state prediction and observation update processes.

[0046] Navigation output module: It calibrates the results of inertial calculation by taking the state errors obtained from the data fusion processing module and outputs the attitude, velocity and position parameters of the calibrated inertial calculation to provide high-precision navigation information for the carrier.

[0047] A GNSS / INS tightly integrated navigation method based on multi-baseline redundant observations, through, as... Figure 1 The integrated navigation device shown is implemented using a three-antenna baseline as an example, as illustrated below:

[0048] Step 1: Coordinate system definition and baseline calibration of the load system

[0049] This method involves four interconnected core coordinate systems. The Earth coordinate system ( The origin of the system is located at Centered on Earth The axis lies in the equatorial plane and points towards the central meridian. The axis is along the direction of the Earth's rotation. , and The three axes satisfy the right-hand rule. (IMU coordinate system) system) origin Located in the sensitive center of the IMU, axis, shaft and The axes are respectively along the horizontal axis to the right, the vertical axis forward, and the vertical axis upward, forming a right-front-upper coordinate system. (Carrier coordinate system) Department): origin Located at the sensitive center of the carrier, axis, shaft and The axes are respectively positioned to the right along the horizontal axis, forward along the vertical axis, and upward along the carrier's vertical axis, forming a right-front-upper coordinate system. (Navigation coordinate system) The navigation coordinate system is the coordinate system selected when calculating navigation parameters. This paper uses the Northeast-Heaven-Earth coordinate system as the navigation coordinate system. In this method design, the IMU coordinate system ( (system) and carrier coordinate system ( IMU coordinate system (IMU coordinate system) installation (system) and carrier coordinate system ( Department) overlap.

[0050] The carrier coordinate system baseline calibration uses a laser tracker or industrial photogrammetry system to perform three-dimensional coordinate measurements of the phase center positions of each antenna in a multi-antenna GNSS system. Further calculations are then performed to determine the carrier coordinate system vectors of the multiple non-collinear short baselines formed between the antennas. These three antenna baselines are denoted as... , , .

[0051] Step 2, Data Acquisition and Calculation

[0052] A multi-antenna GNSS receiver receives signals from five or more satellites, constructs a double-difference observation model, effectively eliminates satellite clock errors and receiver clock errors, and calculates the three baseline vectors corresponding to the calibration baselines in step 1 in the Earth coordinate system. The corresponding vector under the system) , , Simultaneously, the inertial measurement unit of the IMU and inertial calculation module outputs the raw data of the carrier's three-axis angular velocity and specific force in real time. After processing by the strapdown inertial navigation system, the carrier's attitude matrix is ​​calculated. By combining the velocity and position update equations, continuous navigation parameter outputs are generated, including latitude and longitude, altitude, velocity, and Euler attitude angles.

[0053] Step 3, Construction of the state equations for the Kalman filter

[0054] A 15-dimensional state vector is selected, mainly including attitude misalignment angle error. Speed ​​error Position error Gyroscope zero bias error Adding zero offset error State vector It can be represented as:

[0055]

[0056] The continuous-time state transition equation can be expressed as:

[0057]

[0058] In the formula, State variables Time derivative, The state transition matrix is ​​determined by the inertial navigation error propagation relationship. The noise distribution matrix is... This is the system noise vector.

[0059] Step 4, Construction of the Kalman filter observation equation

[0060] This step involves establishing a Kalman filter observation equation based on multi-baseline observations to fuse GNSS baseline observations with INS navigation parameters. Specifically, it includes the following technical steps:

[0061] (1) Calculation of baseline vector for inertial navigation system derivation

[0062] Based on the baseline vector data of the carrier system calibration obtained in step 1, and the real-time attitude matrix output by the IMU in step S2. Complete the baseline vector , , Transformation in the navigation coordinate system. Specifically, for the three baseline vectors of the vehicle system, this is achieved through the attitude transfer matrix. Formula for projecting onto the navigation coordinate system The baseline vector derived by the inertial navigation system is obtained. , , .

[0063] (2) Selection of observations and modeling of observation equations

[0064] Define the baseline vector for GNSS double-difference observation solution. , , With baseline vectors derived from inertial navigation , , In the residuals of the navigation system , , As an observation Its mathematical expression is:

[0065]

[0066] in, The transformation matrix from the Earth coordinate system to the navigation coordinate system can be calculated based on the latitude and longitude information obtained in step 2.

[0067] Furthermore, the relationship between attitude misalignment angle error and baseline vector error is considered. ( (for residual error terms), , , The observation matrix is ​​constructed by stacking three baselines. and residual vector And establish linearized observation equations :

[0068] .

[0069] Step 5, Kalman filter calibration

[0070] The continuous-time state equation and observation equation are discretized and input into a Kalman filter. The attitude misalignment angle error is estimated through a prediction-update process. The estimated error is used to perform real-time calibration and update of the INS attitude output.

[0071] The contents not described in detail in this invention are existing technologies known to those skilled in the art.

Claims

1. A combined navigation device based on multi-baseline redundant observation, characterized in that, It includes a multi-antenna GNSS receiver module, an IMU and inertial calculation module, a data fusion processing module, and a navigation output module; The multi-antenna GNSS receiving module consists of a receiver and at least three GNSS antennas. The GNSS antennas are rigidly connected to the carrier to form multiple non-collinear short baselines, ensuring that the phase center of the GNSS antennas is anchored in the carrier coordinate system. The receiver receives satellite signals and uses carrier phase double difference observation technology and integer period ambiguity fixing method to calculate the baseline vector in the Earth coordinate system. The IMU and inertial calculation module are used to collect the raw data of the carrier's angular velocity and specific force in real time, and to complete the inertial navigation calculation. The data fusion processing module achieves synchronous communication with the multi-antenna GNSS receiving module, IMU, and inertial calculation module through a data interface. It constructs the Kalman filter state equation and Kalman filter observation equation to realize the real-time operation of the Kalman filtering algorithm. Based on the inertial navigation error propagation law, it constructs a 15-dimensional state equation. Based on the baseline error of the GNSS baseline measurement vector and the baseline vector calculated based on the inertial navigation attitude in the navigation coordinate system, it constructs the observation measurement. Using the Kalman filtering algorithm, it iteratively estimates each state error through the state prediction and observation update process. The navigation output module calibrates the inertial calculation results using the state errors obtained from the data fusion processing module, and outputs the calibrated attitude, velocity, and position parameters of the inertial calculation to provide high-precision navigation information for the carrier.

2. A GNSS / INS integrated navigation method based on multi-baseline redundant observation, implemented by the integrated navigation device based on multi-baseline redundant observation as described in claim 1, characterized in that, Specifically, the following steps are included: Step 1: Set the coordinate system and calibrate the baseline of the carrier system, where the coordinate system includes the Earth coordinate system, IMU coordinate system, carrier coordinate system and navigation coordinate system; Step 2: The multi-antenna GNSS receiving module receives signals from more than 5 satellites, constructs a double-difference observation model, eliminates satellite clock errors and receiver clock errors, and calculates the baseline vectors of the carrier system in the Earth coordinate system. , , Simultaneously, the inertial measurement unit of the IMU and inertial calculation module outputs the raw data of the carrier's three-axis angular velocity and specific force in real time. After processing by the strapdown inertial navigation system, the carrier's attitude matrix is ​​calculated. By combining the velocity and position update equations, continuous INS navigation parameter outputs are generated, including latitude and longitude, altitude, velocity, and Euler attitude angles. Step 3: Select a 15-dimensional state vector, including attitude misalignment angle error. Speed ​​error Position error Gyroscope zero bias error and the zero bias error of the table State vector Represented as: ， The continuous-time state transition equation is expressed as: ， In the formula, State variables Time derivative, The state transition matrix is ​​determined by the inertial navigation error propagation relationship. The noise distribution matrix is... This is the system noise vector; Step 4: Combining the relationship between attitude misalignment angle error and baseline vector error, stack the m baselines to construct the observation matrix. and residual vector This leads to the formation of Kalman filter observation equations based on multi-baseline observations; Step 5: Discretize the continuous-time state equation and the Kalman filter observation equation, input them into the Kalman filter, and estimate the attitude misalignment angle error through the prediction-update process. Using attitude misalignment angle error The attitude output of the INS is calibrated and updated in real time.

3. The GNSS / INS integrated navigation method based on multi-baseline redundant observations according to claim 2, characterized in that, The origin of the Earth coordinate system Centered on Earth The axis lies in the equatorial plane and points towards the central meridian. The axis is along the direction of the Earth's rotation. , and The relationship between the three axes satisfies the right-hand rule; IMU coordinate system origin Located in the sensitive center of the IMU, axis, shaft and The axes are respectively aligned with the horizontal axis to the right, the vertical axis forward, and the vertical axis upward, forming a right-front-upper coordinate system; Carrier coordinate system: origin Located at the sensitive center of the carrier, axis, shaft and The axes are respectively positioned to the right along the transverse axis, forward along the longitudinal axis, and upward along the vertical axis of the carrier, forming a right-front-upper coordinate system; Navigation coordinate system: The navigation coordinate system is the coordinate system selected when calculating navigation parameters. The Northeast-Heaven-Earth coordinate system is adopted as the navigation coordinate system. The IMU coordinate system coincides with the carrier coordinate system.

4. The GNSS / INS integrated navigation method based on multi-baseline redundant observations according to claim 2, characterized in that, The carrier coordinate system baseline calibration uses a laser tracker or industrial photogrammetry system to perform three-dimensional coordinate measurements of the phase center positions of each GNSS antenna, and further calculates the carrier coordinate system vectors of the multiple non-collinear short baselines formed between the GNSS antennas. , , And store.

5. A GNSS / INS integrated navigation method based on multi-baseline redundant observations according to claim 2, characterized in that, The specific process of step 4 is as follows: Based on the baseline vector data obtained through the baseline calibration of the carrier system in step 1, and the carrier attitude matrix output by the inertial measurement unit of the IMU and inertial calculation module in step 2. Complete the baseline vector , , Transformation in the navigation coordinate system; For the i-th load system baseline vector It is processed by the attitude transition matrix. Projecting onto the navigation coordinate system yields the baseline vector derived by the inertial navigation system. : ， Define the baseline vector calculated in step 2 and the residual of the baseline vector derived from inertial navigation in the navigation coordinate system as the observation. For the i-th baseline, the baseline vector calculated by GNSS double-difference observation... Baseline vector corresponding to inertial navigation calculation Differential under navigation system That is, observation Its mathematical expression is: ， in, The transformation matrix from the Earth coordinate system to the navigation coordinate system is calculated based on the latitude and longitude information obtained in step 2. Combining the relationship between attitude misalignment angle error and baseline vector error in, The residual error term will include , , An observation matrix is ​​constructed by stacking m baselines, including [the baselines mentioned above]. and residual vector And establish linearized observation equations : 。

Images