An integrated navigation system online calibration method, system, storage medium and terminal

Abstract

The application discloses a kind of combined navigation system online calibration method, system, storage medium and terminal, comprising: establishing the state space model of combined navigation system online calibration, the state space model includes the space installation error of auxiliary sensor, the constant drift of gyroscope, scale factor error and the bias error of accelerometer, while introducing the reference information provided by multiple auxiliary sensors;Error correction is carried out to the calibration parameter and navigation parameter of the online calibration model;The output information of different navigation sensors is information fusion;The state space model is carried out observability analysis and design maneuver trajectory.The present application can solve the problem that part of calibration parameter cannot be estimated and the observability of IMU scale factor error is weak in the existing combined navigation system online calibration method, while improving the calibration efficiency and precision of combined navigation system.

Description

TECHNICAL FIELD

[0001] The present application relates to the online calibration technical field, and particularly relates to a combined navigation system online calibration method, system, storage medium and terminal. BACKGROUND

[0002] Doppler Velocity Log (DVL), Global Navigation Satellite System (GNSS) and Celestial Navigation System (CNS) all have the characteristics that navigation errors do not accumulate with time, but still have the problems of poor dynamic performance and incomplete navigation information. Strap-down Inertial Navigation System (SINS) has the characteristics of good dynamic performance and comprehensive navigation information, but has the problem that navigation errors accumulate with time, so the existing navigation systems usually work in a combined navigation manner. The combined navigation technology mainly combines the performance advantages of different navigation sensors, fuses SINS and other navigation systems, so as to achieve better navigation performance than single navigation sensor.

[0003] The calibration technology of SINS is a process of making the output of SINS device consistent with the known reference information by determining a set of device compensation parameters through comparison of the output of SINS device and the known reference information. The existing mature SINS calibration methods mainly include discrete calibration and system-level calibration, both of which need strong rotation excitation and zero-speed reference information provided by a turntable, have strong dependence on the turntable, and still belong to the internal field calibration method. For SINS installed on a carrier, after the inertial components are used for a period of time, the calibration parameters of the Inertial Measurement Unit (IMU) will change after the aging, maintenance or replacement of the components. In order to ensure the navigation accuracy of SINS, the SINS needs to be recalibrated. The existing internal field calibration method needs to dismount the IMU from the carrier, which brings great difficulty to the calibration of SINS. Therefore, it is urgent to research an online calibration method to avoid the trouble caused by repeated dismounting, and only need to use the maneuverability of the carrier and the maneuverability conditions brought by the external environment to realize the online estimation of the calibration parameters.

[0004] To ensure the safe navigation of the carrier, multiple auxiliary navigation systems such as DVL, GNSS, CNS and the like are usually equipped. For different application scenarios and requirements, SINS / GNSS, SINS / DVL, SINS / CNS and other combined navigation modes can be constituted. If the above combined modes are calibrated one by one online before starting, it will certainly be time-consuming and laborious. In addition, the existing online calibration method of combined navigation has the problems of unobservable partial calibration parameters and weak observability of IMU scale factor error.

[0005] Considering the above problems, combined with the performance advantages of each sensor and the effective maneuver of the carrier, an online calibration method of multi-sensor combined navigation is studied, which is expected to realize the joint online calibration of IMU bias, scale factor error and calibration parameters of each auxiliary sensor of the combined navigation system, and improve the calibration efficiency and accuracy. However, the online calibration of multi-sensor combined navigation system usually has the following problems: (1) limited by the maneuvering conditions of the carrier, the observability of some calibration parameters is unknown, and the observability needs to be explored; (2) the output frequencies of each navigation sensor are different, which makes the information fusion method based on Kalman filter not applicable; (3) the calibration models of each sensor are different, and there are approximation errors introduced by nonlinearity. Therefore, how to design an online calibration method of multi-sensor combined navigation is a problem worth studying. SUMMARY

[0006] The purpose of the present application is to overcome the shortcomings of the existing combined navigation system, and provide an online calibration method, system, storage medium and terminal of a combined navigation system.

[0007] The purpose of the present application is achieved by the following technical solutions:

[0008] In a first aspect, an online calibration method of a combined navigation system is provided, the method comprising:

[0009] Step one, establishing a state space model of online calibration of the combined navigation system, the state space model containing the spatial installation error of the auxiliary sensor, the constant drift of the gyroscope, the scale factor error and the bias error of the accelerometer, and introducing the reference information provided by multiple auxiliary sensors;

[0010] Step two, error correction of the calibration parameters and navigation parameters of the online calibration model;

[0011] Step three, information fusion of the output information of different navigation sensors;

[0012] Step four, observability analysis of the state space model and design of the maneuvering trajectory.

[0013] As a preferred option, a combined navigation system online calibration method, the combined navigation system is a SINS / DVL / GNSS / CNS combined navigation system, wherein the SINS is a main navigation system, and the DVL, GNSS and CNS are auxiliary navigation systems.

[0014] As a preferred option, a combined navigation system online calibration method, the state space model is as follows:

[0015]

[0016] State variables System noise w and w ins are the same,

[0017]

[0018]

[0019]

[0020] wherein ε b is the gyro constant drift, is the accelerometer bias, δk g is the scale factor error of the gyroscope.

[0021] As a preferred option, a combined navigation system online calibration method, an error correction method based on full feedback is used to correct the errors of the calibration parameters and navigation parameters of the online calibration model.

[0022] As a preferred option, a combined navigation system online calibration method, an information fusion strategy based on sequential filtering is used to fuse the outputs of different navigation sensors.

[0023] As a preferred option, a combined navigation system online calibration method, the observability analysis of the state space model includes:

[0024] An observability analysis method based on state mean square error matrix is used to analyze the observability of the selected to-be-estimated parameters.

[0025] As a preferred option, a combined navigation system online calibration method, the observability analysis method based on state mean square error matrix includes:

[0026] The state space model of step one is subjected to 2000s Kalman filtering, and the changes of the covariance matrix in the filtering process are recorded;

[0027] The observability of each state variable is extracted:

[0028]

[0029] wherein P 0|0 is the initial error covariance matrix, P k|k is the kth time state root mean square error matrix, i is the i th state in x;

[0030] In combination with the possible maneuver of the carrier, the observability is tested under the conditions of static base, uniform straight movement, acceleration and deceleration movement, swing base, swing and linear motion combination, and heading change and linear motion combination, and the observability of each state under different maneuvers is obtained.

[0031] The optimal maneuver corresponding to each state is screened, the optimal maneuvers of different states are linearly combined, and a maneuver trajectory for online calibration is designed.

[0032] In a second aspect, an online calibration system of a combined navigation system is provided, and the system comprises:

[0033] A state space model construction module configured to establish a state space model for online calibration of the combined navigation system, wherein the state space model comprises spatial installation errors of auxiliary sensors, constant drifts of gyroscopes, scale factor errors, and bias errors of accelerometers, and reference information provided by multiple auxiliary sensors is introduced;

[0034] An error correction module configured to correct errors of calibration parameters and navigation parameters of the online calibration model;

[0035] An information fusion module configured to fuse outputs of different navigation sensors;

[0036] A maneuver trajectory design module configured to analyze the observability of the state space model and design a maneuver trajectory.

[0037] In a third aspect, a computer storage medium is provided, and the computer storage medium stores computer instructions, and the computer instructions perform the related steps in any one of the online calibration methods of the combined navigation system when running.

[0038] In a fourth aspect, a terminal is provided, and the terminal comprises a memory and a processor, and the memory stores computer instructions executable on the processor, and the processor performs the related steps in any one of the online calibration methods of the combined navigation system when executing the computer instructions.

[0039] It should be further explained that the technical features corresponding to the above options can be combined or replaced with each other to form new technical solutions without conflict.

[0040] Compared with the prior art, the present application has the following advantages:

[0041] (1) The state space model established by the application contains the spatial installation error of the auxiliary sensor, the constant drift and scale factor error of the gyroscope, and the bias error of the accelerometer, introduces multiple auxiliary sensors to provide reference information of position, speed and orientation, and solves the problems of unestimable partial calibration parameters and weak observability of IMU scale factor error in the existing online calibration method of the integrated navigation system.

[0042] (2) The application avoids the problem of decline in calibration parameter estimation accuracy caused by nonlinearity through the error correction scheme based on full feedback, and can still realize online calibration of the scale factor error of the gyroscope in the case of a large installation error angle (within 10 degrees).

[0043] (3) The application adopts the observability analysis method based on the root mean square error of the state to design a maneuvering trajectory, which can ensure that each calibration parameter has strong observability and quickly converges in the calibration process, thereby improving the calibration efficiency and accuracy of the integrated navigation system.

[0044] (4) The online calibration method of the SINS / DVL / GNSS / CNS integrated navigation system provided by the application does not rely on the strong rotating excitation and zero speed reference information provided by the turntable, avoids the trouble caused by disassembling the navigation sensors in the existing field calibration or system-level calibration, reduces the operation burden and cost, and can online jointly calibrate the parameters of the gyroscope constant drift, accelerometer bias, gyroscope scale factor error, GNSS rod arm, DVL rod arm and installation error angle, and CNS azimuth installation error angle. BRIEF DESCRIPTION OF DRAWINGS

[0045] Figure 1 A flowchart of an online calibration method of an integrated navigation system according to an embodiment of the application is shown.

[0046] Figure 2 An error correction framework diagram of online calibration according to an embodiment of the application is shown.

[0047] Figure 3 An update flowchart of a sequential filter according to an embodiment of the application is shown.

[0048] Figure 4 A motion trajectory diagram designed for ship motion according to an embodiment of the application is shown.

[0049] Figure 5 An error curve of the gyroscope constant drift in simulation one according to an embodiment of the application is shown.

[0050] Figure 6 An error curve of the accelerometer bias in simulation one according to an embodiment of the application is shown.

[0051] Figure 7Estimation result of DVL installation error angle in one of the simulations shown by the embodiments of the present application;

[0052] Figure 8 Estimation result of gyro scale factor error in one of the simulations shown by the embodiments of the present application;

[0053] Figure 9 Estimation result of DVL lever arm in one of the simulations shown by the embodiments of the present application;

[0054] Figure 10 Estimation result of GNSS lever arm in one of the simulations shown by the embodiments of the present application;

[0055] Figure 11 Estimation result of CNS azimuth installation error angle in one of the simulations shown by the embodiments of the present application;

[0056] Figure 12 Estimation error of gyro constant drift in the second simulation shown by the embodiments of the present application;

[0057] Figure 13 Estimation error of accelerometer bias in the second simulation shown by the embodiments of the present application;

[0058] Figure 14 Estimation error of gyro scale factor error in the second simulation shown by the embodiments of the present application;

[0059] Figure 15 Estimation error of DVL installation error angle in the second simulation shown by the embodiments of the present application;

[0060] Figure 16 Estimation error of DVL lever arm in the second simulation shown by the embodiments of the present application;

[0061] Figure 17 Estimation error of GNSS lever arm in the second simulation shown by the embodiments of the present application;

[0062] Figure 18 Estimation error of CNS installation error angle in the second simulation shown by the embodiments of the present application. DETAILED DESCRIPTION

[0063] The technical solutions of the present application will be described clearly and completely below with reference to the drawings. Obviously, the described embodiments are some but not all of the embodiments of the present application. Based on the embodiments of the present application, all other embodiments obtained by those skilled in the art without creative work fall within the protection scope of the present application.

[0064] In addition, the technical features involved in different embodiments of the present application described below can be combined with each other as long as there is no conflict.

[0065] Reference Figure 1 In one exemplary embodiment, an online calibration method for a combined navigation system is provided, the method comprising:

[0066] Step 1: Establish a state space model for online calibration of the integrated navigation system. The state space model includes the spatial installation error of auxiliary sensors, the constant drift of gyroscopes, the scaling factor error, and the bias error of accelerometers. Reference information provided by multiple auxiliary sensors is also introduced.

[0067] Step 2: Perform error correction on the calibration parameters and navigation parameters of the online calibration model;

[0068] Step 3: Perform information fusion on the output information of different navigation sensors;

[0069] Step 4: Perform observability analysis on the state-space model and design the maneuver trajectory.

[0070] Step one specifically includes:

[0071] 1. Constructing an error model for SINS devices

[0072] Once the SINS is mounted on the carrier, the mounting error angles of the gyroscope and accelerometer do not change over time. Therefore, the device errors of the gyroscope and accelerometer are only related to the scaling factor error, gyroscope constant drift, accelerometer bias, and random noise of the gyroscope and accelerometer. n, b, e, i, and d represent the navigation coordinate system, carrier coordinate system, Earth coordinate system, inertial coordinate system, and DVL reference coordinate system, respectively.

[0073] (1) Gyroscope output error model

[0074] First, establish the output error of the gyroscope. The model:

[0075]

[0076] In the above formula This represents the projection vector of the angular velocity of the system's rotation relative to the inertial frame onto the carrier. ε represents the angular velocity of the carrier output by the gyroscope. b For the constant drift of the gyroscope, n g For the random noise of the gyroscope, diag[·] denotes converting the column vector[·] into a diagonal matrix, denoted as... As can be seen from the gyroscope output error model, the gyroscope output error is only related to the gyroscope scale factor error parameter δk. g angular velocity of the carrier Gyroscope constant drift ε band gyroscope output random noise n g Related.

[0077] (2) Accelerometer output error model

[0078] This invention uses the accelerometer output error δf b The model is:

[0079]

[0080] For the output of the accelerometer, For comparison, and n a The zero bias of the meter and the random noise of the accelerometer were added respectively. This yielded the device output error model of the SINS, which includes the IMU bias error, scaling factor error, and random noise.

[0081] 2. Construct the SINS navigation error equation.

[0082] The established SINS device error model will be substituted into the SINS navigation error equations to obtain the attitude error equations, velocity error equations, and position error equations required for calibration, which will then be organized into a matrix block form. Attitude error, velocity error, and position error are denoted by the symbol φ. n δv n δp is used to represent, For strapdown matrix, Let L, λ, and h represent latitude, longitude, and altitude, respectively, and δL, δλ, and δh be the corresponding error terms. ie v x v y R represents the Earth's rotational angular rate, eastward velocity, and northward velocity, respectively. m and R n These are the radii of the Mao-You circle and the Zi-Wu circle, respectively.

[0083] (1) Attitude error equation

[0084] Substituting the gyroscope output error model into the attitude error equation and simplifying, we get:

[0085]

[0086] In the above formula:

[0087]

[0088]

[0089]

[0090]

[0091] (2) Velocity error equation

[0092] Substituting the accelerometer output error model into the velocity error equation yields:

[0093]

[0094] in:

[0095]

[0096]

[0097] (3) Position error equation

[0098] Rearranging the position error equation into a matrix block form, we get:

[0099]

[0100] In the above formula:

[0101]

[0102] 3. Construct an auxiliary sensor error model

[0103] As the primary navigation system for the carrier, the SINS (Self-Installing Navigation System) does not need to radiate or receive signals outwards. Therefore, the SINS typically needs to be pre-calibrated (or installed) at the carrier's center of mass. Auxiliary sensors are installed in different locations depending on their measurement principles; for example, GNSS is usually installed at the top of the carrier, while DVL (Digital Volume Level) sensors are usually installed at the bottom. This results in situations where the installation axis and center of mass of the auxiliary sensors do not coincide with the carrier coordinate system of the SINS. This invention refers to these errors as spatial installation errors. The existence of spatial installation errors leads to inaccuracies in the auxiliary information provided by the auxiliary sensors, easily reducing the navigation accuracy of the integrated navigation system. Therefore, the spatial installation errors of the auxiliary sensors need to be calibrated before use. This invention establishes measurement models for DVL, GNSS, and CNS, and simultaneously considers the spatial installation errors of GNSS, DVL, and CNS relative to the SINS. The spatial installation error models for each auxiliary sensor are given below.

[0104] (1) Spatial Installation Error Model of DVL

[0105] Establish DVL output speed The model is:

[0106]

[0107] To form the installation error matrix, the installation error angle θ is usually approximated as a small angle, i.e. Since the relative positions and orientations of the DVL and SINS typically do not change after installation, meaning the DVL's arm length and installation error angle relative to the SINS are constant, then:

[0108]

[0109]

[0110] Projecting the speed calculated by SINS onto the load system Speed ​​compared to the actual output of DVL By taking the difference, we can obtain the measurement equation for the DVL output speed:

[0111]

[0112] (2) Spatial Installation Error Model of GNSS

[0113] GNSS output location and speed Information and the position p and velocity v of SINS n The information satisfies the following relationship:

[0114]

[0115]

[0116] The actual output velocity and position noise of GNSS are n vg and n pg n vg and n pg They follow a pattern with a mean of 0 and variances R0 and R0 respectively vg and R pg White noise. Since the relative positions of the GNSS antenna and SINS are usually unchanged after installation, i.e., the GNSS antenna is positioned relative to the SINS boom arm... If the value is constant under the load system, then:

[0117]

[0118] Speed ​​of SINS calculation and location Speed ​​compared to actual GNSS output and location By subtracting the values, we can obtain the measurement equation for the GNSS output:

[0119]

[0120]

[0121] (3) Spatial Installation Error Model of CNS

[0122] To improve the observability of the SINS gyroscope scaling factor error, the CNS output is introduced as auxiliary information. The error model for the CNS output heading information is as follows:

[0123]

[0124] in θ is the true azimuth angle of SINS. cz Let n be the azimuth installation error angle of CNS relative to SINS. cns The noise output by CNS has a mean of 0 and a variance of R. cns Since the relative positions of SINS and CNS do not change after installation, the azimuth installation error angle of CNS relative to SINS can be established as a constant model:

[0125]

[0126] The azimuth angle calculated by the inertial navigation By subtracting the azimuth angle from the CNS output, the measurement equation for the CNS output azimuth angle can be obtained:

[0127]

[0128] Where φ x φ y and φ z These are the misalignment angle errors in the east, north, and celestial directions, respectively, and φ n =[φ x φ y φ z ] T .

[0129] 4. Construct a state-space model for online calibration of the integrated navigation system.

[0130] (1) System Model

[0131] To control the constant drift ε of the gyroscope b Accelerometer bias gyroscope scaling factor error δk g δk a For calibration, the state-space model established in this invention is as follows:

[0132]

[0133] Where: the state variable is selected as System noise w and wins same,

[0134]

[0135]

[0136]

[0137] (2) Measurement Model

[0138] Considering that the sampling frequencies of GNSS, DVL, and CNS are different, the measurement models for the three auxiliary sensors are established separately. The measurement model for the GNSS output is established as follows:

[0139] z vg =H vg x+n vg

[0140] z pg =H pg x+n pg

[0141] in:

[0142]

[0143] Based on the derived DVL output speed error equation, the measurement model for DVL output is established as follows:

[0144] z dvl =H dvl x+n dvl

[0145] in:

[0146]

[0147] Based on the derived CNS output error equation, the measurement model for CNS output is established as follows:

[0148] z cns =H cns x+n cns

[0149] in:

[0150]

[0151] This completes the establishment of the online calibration model for the SINS / DVL / GNSS / CNS integrated navigation system. The existing state-space model is based on conditions where the misalignment angle, IMU calibration parameters, auxiliary sensor installation error angle, and boom arm are small. Due to the different operating principles of the auxiliary sensors, appropriate installation positions need to be selected based on their operating principles. For example, GNSS antennas are typically installed on top of the carrier to receive radio signals, while DVL antennas are typically installed at the bottom of the carrier to receive acoustic signals returning from the bottom. Therefore, larger boom arm and installation error angles may occur. Large spatial installation errors can easily introduce nonlinearity into the measurement model, leading to inaccuracies and affecting the online calibration accuracy. Therefore, step two is needed to avoid the decrease in filtering accuracy caused by nonlinearity.

[0152] Although the established online calibration model includes models for IMU scaling factor error and DVL lever arm error, the observability of calibration parameters such as IMU scaling factor error and DVL lever arm error remains unknown due to limitations in the maneuverability of the carrier. Therefore, it is necessary to perform step four, under the maneuverability trajectory achievable by the carrier, to conduct an observability analysis of the established SINS / DVL / GNSS / CNS integrated navigation model and determine effective maneuvers, providing a basis for the design of the motion trajectory for online carrier calibration.

[0153] Preferably, to address the problem of nonlinear errors caused by large calibration parameters, this invention designs an online calibration error correction scheme based on full feedback, overcoming the performance degradation of Kalman filter estimation caused by nonlinearity of calibration parameters. Specific details include:

[0154] As can be seen from the established state-space model, the spatial installation error of the auxiliary sensor is unknown, which may lead to a large spatial installation error, thus introducing model nonlinearity and resulting in approximation errors. To avoid the problem of decreased accuracy in calibration parameter estimation caused by nonlinearity, this invention employs a full feedback correction method to correct the errors in the calibration and navigation parameters of the online calibration model. The block diagram of the full feedback correction principle proposed in step two is as follows: Figure 2 As shown. From Figure 2 As can be seen, the error correction block diagram of online calibration contains two error correction loops: one is the navigation parameter feedback correction loop, and the other is the navigation sensor calibration parameter error correction loop.

[0155] The specific details of feedback correction in the online calibration model are as follows:

[0156] (1) Navigation parameter error correction

[0157] By directly compensating for navigation error estimates in navigation parameters, feedback correction of navigation error-related state variables can be achieved, i.e.:

[0158]

[0159] in and For t k The attitude matrix, velocity, and position are constantly updated through inertial navigation system calculations. and These are the attitude matrix, velocity, and position after navigation error correction. Feedback correction of navigation error-related state variables involves adjusting the navigation parameters after error correction. and This serves as the initial value for the inertial navigation system's solution update at the next moment.

[0160] (2) Calibration parameter error correction

[0161] For calibration parameter-type state variables, the estimated calibration parameter error is first compensated into the calibration parameters, and then the compensated calibration parameters are fed back to the output of each sensor. Therefore, the Kalman filter in the feedback correction estimates the residual error of each calibration parameter. This invention defines the errors in the feedback correction—gyroscope constant drift, accelerometer bias, gyroscope scaling factor error, DVL lever arm and installation error angle, and GNSS lever arm and CNS installation error angle—as follows: The relationship between the calibration parameter error estimated at time k and the calibration parameter estimated at time k-1 is as follows:

[0162]

[0163] in These are the calibration parameters estimated at the previous time step. These are the calibration parameters after compensation at the current moment. The estimated calibration parameters also need to be fed back to the output error model of each sensor in real time. The specific feedback compensation process is as follows:

[0164]

[0165] in and These are the compensated gyroscope output, accelerometer output, DVL output, GNSS output, and CNS output, respectively. The compensated gyroscope output... Inertial navigation solution obtained Angular velocity can be obtained

[0166]

[0167] Based on the latest estimated installation error angle The compensated installation error matrix can be obtained.

[0168] Once the calibration parameter error estimated by the Kalman filter has been compensated into the calibration parameters, the residual error of the calibration parameters estimated at the current time can be considered to be zero. Therefore, the filter state corresponding to the next time step needs to be set to zero.

[0169] Preferably, to address the issue of different output frequencies of various navigation sensors, this invention achieves information fusion of the integrated navigation system at different frequencies through an information fusion strategy based on sequential filtering, specifically including:

[0170] The output frequencies of SINS, DVL, GNSS, and CNS cannot be guaranteed to be completely consistent. Usually, the output frequency of SINS is much higher than that of the auxiliary sensors. Therefore, it is necessary to select an appropriate information fusion method to achieve online calibration of the integrated navigation system.

[0171] This invention employs a sequential filter for information fusion, which can divide measurement updates into N sub-measurements for updating. For all sub-measurements arriving at time k, measurement updates are performed using recursive least squares estimation, and the update flowchart is as follows. Figure 3 As shown.

[0172] Preferably, to address the issue of unknown observability of some calibration parameters, an observability analysis method based on the root mean square error matrix of the state is used to perform observability analysis on the selected parameters to be estimated. Based on the observability analysis, the maneuver trajectory is designed, specifically including:

[0173] Perform a 2000s Kalman filter on the online calibration state-space model described in step one, and record the changes in the covariance matrix during the filtering process.

[0174] An observability analysis method based on root mean square error of state is used to extract the observability of each state variable:

[0175]

[0176] Where P 0|0 Let P be the initial error covariance matrix. k|k Let be the root mean square error matrix of the state at time k, and let i be the i-th state in x;

[0177] Based on the possible maneuvers of the carrier, observability tests were conducted under different maneuvers, including static base, uniform straight-line motion, acceleration and deceleration motion, swaying base, combination of swaying and linear motion, and combination of heading change and linear motion, to obtain the observability of each state under different maneuvers.

[0178] The optimal maneuver for each state is selected, and the optimal maneuvers for different states are linearly combined to design an online calibrated maneuver trajectory.

[0179] Since the maneuver trajectory includes a variety of maneuver forms, different combinations and movement durations can design different maneuver trajectories. Therefore, this invention only provides a design idea for online calibration of maneuver trajectory design. In practical applications, more reasonable maneuver trajectories can be designed according to specific situations and needs.

[0180] In another exemplary embodiment, the above method is illustrated in conjunction with a practical application scenario. To ensure safe navigation of ships, multiple auxiliary navigation systems are typically equipped, such as DVL, GNSS, and CNS. GNSS can provide speed and position information, DVL can provide speed correction information for SINS in the event of GNSS rejection, and CNS can provide high-precision heading correction information for SINS in good weather conditions. Depending on different application scenarios and requirements, various combined navigation modes such as SINS / GNSS, SINS / DVL, and SINS / CNS can be configured. Performing online calibration for each of these combined modes before startup would be time-consuming and labor-intensive. Existing online calibration methods for combined navigation systems focus on calibrating the spatial installation error of the auxiliary sensors relative to the IMU, without fully utilizing the available auxiliary navigation information on the ship. This results in problems such as the inability to estimate some calibration parameters and weak observability of IMU scaling factor errors.

[0181] The states of a marine SINS / DVL / GNSS / CNS integrated navigation system can be categorized into navigation error states and calibration parameter states. Attitude error, velocity error, and position error are navigation error state variables, while residual gyroscope constant drift, accelerometer bias, DVL lever and installation error angles, and GNSS lever and CNS installation error angles are calibration parameter state variables. See step two for specific calibration details.

[0182] Next, the sequential filtering-based information fusion strategy described in step three is used to achieve information fusion of the integrated navigation system at different frequencies, while avoiding the performance degradation of Kalman filter estimation caused by the nonlinearity of calibration parameters.

[0183] Then, addressing the issue of unknown observability of some calibration parameters, the observability analysis method based on the root mean square error matrix of the state, as described in step four, was used to perform observability analysis on the selected parameters to be estimated. Based on the observability analysis results, the observable maneuvers and optimal observable maneuvers for each state were summarized, as shown in Table 1. In Table 1, the six motion modes—static base, uniform motion, acceleration / deceleration motion, rocking base, a combination of rocking and uniform motion, and a combination of turning and uniform motion—are labeled as Motion 1 to Motion 6, respectively.

[0184] Table 1. Observability test results for each calibration parameter

[0185]

[0186]

[0187] As shown in Table 1, each calibration parameter achieved the strongest observability in motion states 3 through 6. To ensure that each state can be quickly excited, several sets of acceleration / deceleration motions and heading / turning motions can be added to the combination of the swaying base and linear motion, so that all states are excited by the corresponding optimal maneuver. Based on this, this embodiment designs the online calibration motion trajectory as shown in Table 2 according to the motion sequence that is easier for the ship to achieve, and superimposes the swaying motion on it.

[0188] Table 2. Detailed motion information of the maneuver trajectory

[0189]

[0190]

[0191] The linear and angular velocities of the ship's motion in Table 2 are set based on empirical values ​​summarized from sea trials. The linear and angular velocities in the designed maneuver trajectories are all less than or equal to 10 m / s and 2° / s, respectively, which meets the maneuverability requirements for ships navigating at sea. The designed motion trajectory is as follows: Figure 4 As shown, the designed motion trajectories consist of simple uniform motion, acceleration / deceleration motion, swaying motion, and turning motion, encompassing the optimal maneuvers required for observing all states, and meeting the maneuver trajectory requirements for online calibration of the SINS / DVL / GNSS / CNS integrated navigation system.

[0192] Finally, simulation experiments verified the effectiveness and correctness of the proposed online calibration method for the SINS / DVL / GNSS / CNS integrated navigation system. The simulation compared the following three schemes:

[0193] Option 1: Perform online calibration of the existing SINS / DVL integrated navigation model. The existing SINS / DVL integrated navigation model only models calibration parameters such as gyroscope constant drift, accelerometer bias, and DVL installation error angle, without considering gyroscope scaling factor error and DVL lever error.

[0194] Option 2: Perform online calibration on the model described in step one of this invention, and use partial feedback to correct and compensate for errors. Select attitude, velocity, and position for feedback correction, and perform output correction only for other states.

[0195] Option 3: Perform online calibration on the model described in step one of this invention, and use the full feedback error correction scheme described in step two to perform error correction and compensation, and perform feedback correction on all state parameters.

[0196] In this simulation experiment, the output frequencies of SINS, DVL, GNSS, and CNS were set to 100Hz, 1Hz, 10Hz, and 0.1Hz, respectively. Therefore, a sequential filter was used for information fusion, and the feedback correction frequency was the same as the lowest output frequency of the auxiliary sensor. The initial position of the motion trajectory was: latitude 45.78°, longitude 126.67°, initial velocity 0m / s, and initial attitude [0°; 0°; -30°]. In practice, initial alignment can ensure the attitude error angle φ. n Since the angle is small, but the installation error angle and lever arm are unknown, there may be cases with large installation error angles and lever arms. Therefore, this embodiment conducted two sets of online calibration simulation experiments with different degrees of nonlinearity. Simulation 1 is the online calibration test under small installation error angles and lever arms, and Simulation 2 is the online calibration test under large installation error angles and lever arms.

[0197] The IMU accuracy parameters are set as shown in Table 3, and the auxiliary sensor accuracy parameters are set as shown in Table 4.

[0198] Table 3. IMU accuracy parameter settings in simulation.

[0199]

[0200] Table 4. Accuracy Settings for Auxiliary Sensors

[0201]

[0202] Figures 5-7 The online calibration results for gyroscope constant drift, accelerometer zero bias, and DVL installation error angle in Simulation 1 are shown. The black dashed line represents the online calibration results for Scheme 1, the blue dotted line represents the online calibration results for Scheme 2, and the red solid line represents the online calibration results for Scheme 3.

[0203] from Figures 5-7 It can be seen that, under small installation errors and lever arm conditions, Scheme 2 and Scheme 3 show almost no difference in the estimation performance of IMU calibration parameters and DVL installation error angles. However, Scheme 1 has lower accuracy in estimating IMU calibration parameters and DVL installation error angles, and it cannot estimate the z-axis gyroscope constant drift. The experimental results from Simulation 1 show that, compared to Scheme 1, Schemes 2 and 3 consider gyroscope scaling factor error and DVL lever arm error, and introduce GNSS and CNS to provide velocity, position, and azimuth information references, thus improving the observability and estimation accuracy of IMU calibration parameters and DVL installation error angles.

[0204] from Figures 8-11It can be seen that, under small installation errors and lever arm conditions, both Scheme 2 and Scheme 3 can estimate the gyroscope scaling factor error, DVL lever arm error, GNSS lever arm error, and CNS azimuth installation error angle. Among them, Scheme 3 estimates the DVL lever arm with slightly higher accuracy than Scheme 4. Simulation 1 of the embodiment demonstrates the correctness and effectiveness of the online calibration model established in step 1 of the present invention, and also verifies the effectiveness of the designed maneuver trajectory.

[0205] Furthermore, in order to demonstrate the beneficial effects of the error correction scheme based on full feedback designed in step two of this invention, the IMU parameters and auxiliary sensor parameters in simulation two are set as shown in Tables 3 and 4. The online calibration effect under the conditions of large installation error angle and lever arm is verified by comparing scheme two and scheme three. Figures 12-18 The results are from simulation 2.

[0206] from Figures 12-18 It can be seen that the online calibration accuracy of the GNSS boom in Scheme 2 and Scheme 3 is comparable, both converging quickly to within 0.01m. The online calibration accuracy of gyroscope constant drift, accelerometer bias, and CNS installation error angle in Scheme 2 is slightly lower than that in Scheme 3, but it still achieves high calibration accuracy within 2 hours. Specifically, the gyroscope constant converges to within 0.001° / h, the accelerometer bias to within 2ug, and the CNS installation error angle to within 0.01°. The accuracy of the gyroscope scaling factor error estimated by Scheme 2 is significantly lower than that of Scheme 3, with Scheme 2 exhibiting a large deviation, while Scheme 3's estimated gyroscope scaling factor errors all converge to within 15ppm. The online calibration accuracy of the DVL installation error angle and boom in Scheme 2 is significantly lower than that in Scheme 3, and the calibration results of the estimated installation error angle and boom in Scheme 2 have large residual errors. Specifically, the DVL installation error angle θ estimated by Scheme 2... x θ y and θ z The residual error angles are above 0.04°, 0.09°, and 0.39°, respectively, while the three installation error angles θ estimated by Scheme 3 are higher. x θ y and θ z All converged to within 0.04°, specifically to 0.02°, 0.03°, and 0.01° respectively. The residual errors of the DVL's x, y, and z axis lever arms estimated by Scheme 2 were above 0.9m, 0.3m, and 0.02m respectively, while the lever arms of all three axes of the DVL estimated by Scheme 3 converged to within 0.02m. This is mainly because Scheme 3 adopted the full-feedback-based error correction scheme described in step two of this invention, gradually compressing the estimated lever arm errors and residual installation error angles to a small amount, thereby suppressing the influence of nonlinear errors.

[0207] The experimental results from both Simulation 1 and Simulation 2 show that the maneuvering trajectory designed according to step four of this invention can effectively excite gyroscope constant drift, accelerometer bias, gyroscope scaling factor error, GNSS lever arm, DVL lever arm and installation error angle, and CNS azimuth installation error angle. This indicates that the designed maneuvering trajectory is effective, and the designed trajectory is based on the possible motion states of the ship. Therefore, this trajectory is suitable for online calibration of marine SINS / DVL / GNSS / CNS integrated navigation systems. The simulation results also show that partial feedback of attitude, velocity, and position can achieve online calibration under small installation errors and lever arm conditions, and the calibration results are comparable to those of full-parameter feedback. For cases where large auxiliary sensor calibration parameters cause nonlinearity, the full-parameter feedback correction scheme can achieve online calibration under large installation errors and lever arm conditions, and the calibration accuracy is comparable to that under small installation errors and lever arm conditions. Considering the possibility that the auxiliary sensor calibration parameters of a marine SINS / DVL / GNSS / CNS integrated navigation system may be large and cause nonlinearity, without loss of generality, the online calibration method for the SINS / DVL / GNSS / CNS integrated navigation system designed in steps one to four of this invention can achieve better calibration results.

[0208] Simulation results show that the online calibration method for the SINS / DVL / GNSS / CNS integrated navigation system proposed in this invention can suppress the influence of nonlinearity in calibration parameters and achieve joint online calibration of gyroscope constant drift, accelerometer bias, gyroscope scaling factor error, GNSS lever arm, DVL lever arm and installation error angle, and CNS azimuth installation error angle, thus improving the calibration efficiency and accuracy of marine integrated navigation systems. If the integrated navigation system has more navigation sensors, the design concept of the online calibration method of this invention can also be adopted to improve the online calibration efficiency of integrated navigation systems.

[0209] In another exemplary embodiment, an online calibration system for a combined navigation system is provided, the system comprising:

[0210] The state space model construction module is configured to establish a state space model for online calibration of the integrated navigation system. The state space model includes the spatial installation error of the auxiliary sensors, the constant drift of the gyroscope, the scaling factor error, and the bias error of the accelerometer, while also incorporating reference information provided by multiple auxiliary sensors.

[0211] The error correction module is configured to perform error correction on the calibration parameters and navigation parameters of the online calibration model;

[0212] The information fusion module is configured to fuse the output information from different navigation sensors.

[0213] The maneuver trajectory design module is configured to perform observability analysis on the state space model and design maneuver trajectories.

[0214] In another exemplary embodiment, the present invention provides a computer storage medium storing computer instructions thereon, which, when executed, perform the relevant steps of the online calibration method for a combined navigation system.

[0215] Based on this understanding, the technical solution of this embodiment, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods of the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0216] In another exemplary embodiment, the present invention provides a terminal including a memory and a processor, wherein the memory stores computer instructions that can be executed on the processor, and the processor executes the relevant steps of the online calibration method for the integrated navigation system when executing the computer instructions.

[0217] The processor may be a single-core or multi-core central processing unit or a specific integrated circuit, or one or more integrated circuits configured to implement the present invention.

[0218] Suitable processors for executing computer programs include, for example, general-purpose and / or special-purpose microprocessors, or any other type of central processing unit. Typically, the central processing unit receives instructions and data from read-only memory and / or random access memory. The basic components of a computer include a central processing unit for implementing or executing instructions and one or more memory devices for storing instructions and data. Typically, a computer will also include one or more mass storage devices for storing data, such as disks, magneto-optical disks, or optical disks, or the computer will be operatively coupled to such mass storage devices to receive data from or transfer data to them, or both. However, a computer is not required to have such devices. Furthermore, a computer can be embedded in another device, such as a mobile phone, a personal digital assistant (PDA), a mobile audio or video player, a game console, a global positioning system (GPS) receiver, or a portable storage device such as a universal serial bus (USB) flash drive, to name a few.

[0219] While this specification contains numerous specific implementation details, these should not be construed as limiting the scope of any invention or the scope of the claims, but rather are primarily intended to describe features of specific embodiments of a particular invention. Certain features described in the various embodiments herein may also be implemented in combination in a single embodiment. Conversely, various features described in a single embodiment may also be implemented separately in various embodiments or in any suitable sub-combination. Furthermore, while features may function in certain combinations as described above and even initially claimed in this way, one or more features from a claimed combination may be removed from that combination in some cases, and a claimed combination may refer to a sub-combination or a variation thereof.

[0220] Similarly, although the operations are depicted in a specific order in the accompanying drawings, this should not be construed as requiring these operations to be performed in the specific order shown or sequentially, or requiring all illustrated operations to be performed to achieve the desired result. In some cases, multitasking and parallel processing may be advantageous. Furthermore, the separation of various system modules and components in the above embodiments should not be construed as requiring such separation in all embodiments, and it should be understood that the described program components and systems can generally be integrated together in a single software product or packaged into multiple software products.

[0221] The above detailed embodiments are a description of the present invention. It should not be considered that the specific embodiments of the present invention are limited to these descriptions. For those skilled in the art, several simple deductions and substitutions can be made without departing from the concept of the present invention, and all of these should be considered to fall within the protection scope of the present invention.

Claims

1. An online calibration method for a combined navigation system, characterized in that, The method includes: Step 1: Establish a state-space model for the online calibration of the integrated navigation system. This state-space model includes the spatial installation errors of auxiliary sensors, the constant drift of the gyroscope, the scaling factor error, and the bias error of the accelerometer. It also incorporates reference information provided by multiple auxiliary sensors. The integrated navigation system is a SINS / DVL / GNSS / CNS integrated navigation system, where SINS represents the strapdown inertial navigation system as the primary navigation system, and DVL, GNSS, and CNS are all auxiliary navigation systems. DVL represents a Doppler log, GNSS represents a global satellite navigation system, and CNS represents a celestial navigation system. The state-space model is as follows: State variables System noise and same, , , , , , ,in, For gyroscope constant drift, For accelerometer bias, This refers to the scaling factor error of the gyroscope. This represents the attitude error vector under the main navigation system. This represents the velocity error vector under the main navigation system. This represents the position error vector under the main navigation system. This indicates the installation position deviation of DVL relative to SINS. This indicates the installation error angle of DVL relative to SINS; Indicates the lever arm of GNSS relative to SINS; The azimuth installation error angle of CNS relative to SINS; Step 2: Perform error correction on the calibration parameters and navigation parameters of the online calibration model; Step 3: Perform information fusion on the output information of different navigation sensors; Step 4: Perform observability analysis on the state-space model and design the maneuver trajectory.

2. The online calibration method for a combined navigation system according to claim 1, characterized in that, An error correction method based on full feedback is used to correct the errors in the calibration parameters and navigation parameters of the online calibration model.

3. The online calibration method for a combined navigation system according to claim 1, characterized in that, An information fusion strategy based on sequential filtering is adopted to fuse the outputs of different navigation sensors.

4. The online calibration method for a combined navigation system according to claim 1, characterized in that, The observability analysis of the state-space model includes: An observability analysis method based on the root mean square error matrix of the state is used to perform observability analysis on the selected parameters to be estimated.

5. The online calibration method for a combined navigation system according to claim 4, characterized in that, The observability analysis method based on the root mean square error matrix of the state includes: Perform a 2000s Kalman filter on the state-space model described in step one and record the changes in the covariance matrix during the filtering process. Extract the observability of each state variable: ,in Let the initial error covariance matrix be... For the first The root mean square error matrix of the state at each time step for The first in One state; Based on the possible maneuvers of the carrier, observability tests were conducted under different maneuvers, including static base, uniform straight-line motion, acceleration and deceleration motion, swaying base, combination of swaying and linear motion, and combination of heading change and linear motion, to obtain the observability of each state under different maneuvers. The optimal maneuver for each state is selected, and the optimal maneuvers for different states are linearly combined to design an online calibrated maneuver trajectory.

6. An online calibration system for a combined navigation system, characterized in that, The system includes: The state-space model construction module is configured to establish a state-space model for online calibration of the integrated navigation system. This state-space model includes the spatial installation errors of auxiliary sensors, the constant drift of the gyroscope, the scaling factor error, and the bias error of the accelerometer. It also incorporates reference information provided by multiple auxiliary sensors. The integrated navigation system is a SINS / DVL / GNSS / CNS integrated navigation system, where SINS represents the strapdown inertial navigation system as the primary navigation system, and DVL, GNSS, and CNS are all auxiliary navigation systems. DVL represents a Doppler log, GNSS represents a global satellite navigation system, and CNS represents a celestial navigation system. The state-space model is as follows: State variables System noise and same, , , , , , ,in, For gyroscope constant drift, For accelerometer bias, This refers to the scaling factor error of the gyroscope. This represents the attitude error vector under the main navigation system. This represents the velocity error vector under the main navigation system. This represents the position error vector under the main navigation system. This indicates the installation position deviation of DVL relative to SINS. This indicates the installation error angle of DVL relative to SINS; Indicates the lever arm of GNSS relative to SINS; The azimuth installation error angle of CNS relative to SINS; The error correction module is configured to perform error correction on the calibration parameters and navigation parameters of the online calibration model; The information fusion module is configured to fuse the output information from different navigation sensors. The maneuver trajectory design module is configured to perform observability analysis on the state space model and design maneuver trajectories.

7. A computer storage medium storing computer instructions thereon, characterized in that, When the computer instructions are executed, they perform the relevant steps in the online calibration method for a combined navigation system as described in any one of claims 1-5.

8. A terminal, comprising a memory and a processor, wherein the memory stores computer instructions executable by the processor, characterized in that, When the processor executes computer instructions, it performs the relevant steps in the online calibration method for an integrated navigation system as described in any one of claims 1-5.

Images