Sneaking and crawling combined navigation method and navigation system for amphibious robot

Through the combination of error state Kalman filter and Li Qun Li algebra method, high-precision positioning of amphibious robots in stealth and crawling modes is achieved, and the problem of insufficient positioning accuracy in multi-environment and multi-mode in the prior art is solved. Continuous and reliable navigation solutions in complex underwater environments.

CN118913286APending Publication Date: 2024-11-08NANJING UNIV OF SCI & TECH
View PDF 0 Cites 5 Cited by

Patent Information

Application Number
CN202411032824.7
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-07-30
Publication Date
2024-11-08

AI Technical Summary

Technical Problem

The existing amphibious robot navigation technology is difficult to achieve accurate fusion positioning in multiple environments and multi-modes, especially when switching between stealth and crawling modes, which limits its application effect in complex terrain. .

Method used

Using error state Kalman filter (ESKF) combined with Liqun Li algebra method, the rotation matrix is ​​mapped to the rotation vector, state transfer equations and observation equations are established, and data fusion is carried out through Doppler velocity meter and odometer data to achieve different Fusion positioning under working conditions.

Benefits of technology

It significantly improves the positioning accuracy and stability of underwater robots in different motion modes, provides a continuous and reliable navigation solution in complex underwater environments, and improves practical application effects and task execution capabilities.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118913286A_ABST
    Figure CN118913286A_ABST
Patent Text Reader

Abstract

The invention relates to a sneaking and crawling combined navigation method and navigation system for an amphibious robot. The method comprises the following steps: (1) initially aligning the navigation system; (2) establishing a state transition equation: mapping the rotation matrix to a rotation vector based on a Lie group Lie algebra method, performing track plotting on an inertial measurement unit (IMU), and establishing the state transition equation; (3) establishing an observation equation: establishing an observation equation taking a Doppler velocity meter DVL as an observation source and an observation equation taking an odometer OD as an observation source; and (4) data fusion: carrying out a prediction process of an error state Kalman filter by using a state transition equation, and carrying out an updating process of the error state Kalman filter through an observation equation to realize fusion positioning under different working conditions. According to the method, a continuous and reliable navigation solution can be provided in various complex underwater environments, the positioning precision and stability of the underwater robot in different underwater motion modes can be remarkably improved, and the method is suitable for the fields of ocean detection, environment monitoring and the like.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The invention belongs to the field of navigation, and in particular relates to a stealth and crawling combined navigation method and a navigation system for an amphibious robot. Background Art

[0002] With the rapid development of science and technology, robotics technology is increasingly used in various fields. Especially in exploration and operation tasks in complex environments, traditional single-mode robots often find it difficult to cope with diverse terrain requirements. Therefore, amphibious robots with multiple modes of movement have emerged. Amphibious robots can move flexibly on land and in water, greatly expanding the scope of application of robots. However, in practical applications, how to achieve efficient positioning and navigation of amphibious robots in different environments has become a technical problem that needs to be solved urgently.

[0003] Existing amphibious robot navigation technology mainly focuses on positioning systems in simple environments, and research on how to achieve accurate fusion positioning in multiple environments is still relatively weak. Especially when the robot needs to frequently switch between stealth and crawling modes, how to ensure the accuracy and stability of its positioning is a technical challenge. In addition, the shortcomings of existing technologies in positioning accuracy in complex terrains and multi-sensor data fusion limit the application effect of amphibious robots in actual tasks. Summary of the invention

[0004] The purpose of the present invention is to provide a combined navigation method for an underwater robot that can dive and crawl, thereby realizing the posture prediction of the underwater robot in different navigation states (dive / crawling) and improving the positioning accuracy of this type of underwater robot during underwater operations.

[0005] The technical solution to achieve the purpose of the present invention is: a stealth and crawling combined navigation method of an amphibious robot, comprising the following steps:

[0006] Step (1): Initial alignment of the navigation system: Determine the conversion relationship between the carrier inertial coordinate system and the navigation coordinate system through the accelerometer and gyroscope to obtain the initial attitude matrix;

[0007] Step (2): Establishing the state transfer equation: Mapping the rotation matrix to the rotation vector based on the Lie group Lie algebra method and performing dead reckoning on the inertial measurement unit IMU to establish the state transfer equation;

[0008] Step (3): Establish observation equations: Establish observation equations with the Doppler velocity meter DVL as the observation source and observation equations with the odometer OD as the observation source respectively;

[0009] Step (4): Data fusion: Use the state transfer equation established in step (2) to perform the prediction process of the error state Kalman filter ESKF, and use the observation equation established in step (3) to perform the update process of the error state Kalman filter ESKF to achieve fusion positioning under different working conditions.

[0010] Furthermore, step (1) includes the following steps:

[0011] Step (11): Decomposing the initial posture matrix to be solved;

[0012] Step (12): gradually solve each sub-matrix of the decomposed initial posture matrix to obtain the initial posture matrix and complete the initial alignment.

[0013] Furthermore, step (11) is specifically as follows:

[0014] At the initial moment of alignment, the carrier inertial coordinate system b0 and the navigation coordinate system n0 are defined. The carrier inertial coordinate system b0 coincides with the carrier inertial coordinate system b at the start of alignment, and then does not rotate relative to the inertial space. The n0 system coincides with the navigation coordinate system (n system) at the start of alignment, and then does not rotate relative to the inertial space. Initial attitude matrix The overall breakdown is as follows:

[0015]

[0016] Furthermore, step (12) is specifically as follows:

[0017] For the submatrix

[0018]

[0019] In the formula, is the angular velocity vector of the Earth's rotation, which means the fixed-axis rotation of the n system relative to the n0 system

[0020]

[0021] ω ie represents the magnitude of the Earth's rotation angular velocity, and L represents the local latitude;

[0022] Submatrix The calculation method in the time period t is as follows:

[0023]

[0024] Where L represents the local latitude;

[0025] for

[0026]

[0027] In the formula, Represents gyroscope measurement value, attitude matrix The initial value of is set to the unit matrix I;

[0028] Pair Matrix Solve it and get from formula (5) back, The projection under the b0 system is

[0029]

[0030] From formula (4), we can get Then, we get the projection of the gravity vector in the n0 system.

[0031]

[0032] Integrate equations (6) and (7) and record:

[0033]

[0034] In the formula, i = 1, 2, t2 = 2t1, t2 is the end time of initial alignment, and the relationship between the inertial navigation specific force and the gravity vector is given as:

[0035]

[0036] The superscript ~ indicates the actual measured value. It represents the accelerometer measurement error and linear acceleration disturbance in the b0 system;

[0037] Integrate equation (9) and substitute equation (8) into it to obtain:

[0038]

[0039] neglect The interference of the double vector method is used to determine the attitude, and we get The final result

[0040] Furthermore, step (2) specifically includes the following steps:

[0041] Step (21): Mapping relationship from rotation matrix to rotation vector:

[0042] According to BCH approximation:

[0043]

[0044] On the contrary:

[0045] exp((Φ+△Φ) ∧ )=exp((J l △Φ) ∧ )exp(Φ ∧ )=exp(Φ ∧ )exp((J r △Φ) ∧ ) (13)

[0046] in:

[0047]

[0048]

[0049] J r (φ)=J l (-φ) (16)

[0050] Step (22): Set the state variables and use the manifold method to find the Jacobian matrix F of the error state variables:

[0051] Set the state variable x = [p, v, R, b g ,b a ,g] T , where p is translation, v is velocity, R is rotation matrix, b g ,b a is the gyroscope and the table zero bias, g is the gravity;

[0052] Discrete-time equation of motion for the error state variable:

[0053]

[0054] The corresponding Jacobian matrix F is:

[0055]

[0056] Furthermore, step (3) specifically includes the following steps:

[0057] Step (31): Establish the observation equation with DVL as the observation source:

[0058] The DVL observation of velocity is taken as the observation of velocity in the x, y, and z axes of the carrier system, then:

[0059] The velocity observation in vector form in the navigation coordinate system is written as:

[0060] v dvl =[v x ,v y ,v z ] (19)

[0061] where v x ,v y ,v z Represents the three-axis velocity components in the navigation coordinate system.

[0062] The corresponding observation model is:

[0063] z dvl =v+δv+ω dvl (20)

[0064] Where v represents the nominal state variable, δv represents the error state variable, ω dvl represents DVL observation noise; observation model Z dvl The Jacobian matrix for the error variable δx

[0065] Step (32): Establish the observation equation with the odometer as the observation source:

[0066] Taking the upper left coordinate system, the wheel speed observation in vector form is recorded as:

[0067] v od =[v od ,0,0] T (twenty one)

[0068] The corresponding observation model is:

[0069] v od =R T v (22)

[0070] Among them, R,v is the current state of the vehicle. According to the current estimated R, v od Switch to the navigation coordinate system and directly regard it as the observation of v:

[0071] R od =v (23)

[0072] The corresponding observation model is:

[0073] z od =v+δv+ω od (twenty four)

[0074] Where v represents the nominal state variable, δv represents the error state variable, ω od represents the odometer observation noise;

[0075] Observation Model Z od The Jacobian matrix for the error variable δx

[0076]

[0077] Further, step (4) includes the following steps:

[0078] Step (41): Establish the prediction equation of the state variables, including the prediction of the nominal state variables and the error state variables.

[0079] Prediction of state variables:

[0080]

[0081]

[0082] Step (42): Update the ESKF according to the observation equation corresponding to the crawling or sneaking condition:

[0083] While sneaking:

[0084]

[0085] While crawling:

[0086]

[0087] Step (43) combines the error state with the nominal state to obtain the final fused positioning data:

[0088] After prediction and updating, the error state estimate is corrected, the error state is classified as the nominal state, and then the ESKF is reset. The final integrated navigation positioning information is expressed as:

[0089] p k+1 =p k +δp k

[0090] v k+1 =v k +δv k

[0091] R k+1 =R k Exp(δθ k )

[0092] b g,k+1 =b g,k +δb g,k (29).

[0093] b a,k+1 =b a,k +δb a,k

[0094] g k+1 =g k +δg k

[0095] A stealth and crawling combined navigation system for an amphibious robot adopts the above-mentioned navigation method for navigation.

[0096] Further, the system includes a Doppler acoustic Doppler velocity meter DVL, an inertial measurement unit IMU, an odometer OD and an error state Kalman filter;

[0097] The error state Kalman filter has an observation model Z with the Doppler acoustic Doppler velocity meter DVL as the observation source dvl And the observation model Z with odometer OD as the observation source od ;

[0098] In stealth mode, the Doppler acoustic Doppler velocimeter DVL is used to obtain velocity information, combined with the inertial measurement unit IMU data, and the error state Kalman filter uses the observation model Z dvl ; In the crawling mode, the odometer OD is used to obtain motion information, combined with the IMU data, and the error state Kalman filter uses the observation model Z od .

[0099] Compared with the prior art, the present invention has the following significant advantages:

[0100] The present invention uses the error state Kalman filter (ESKF) and the Lie group algebra method to map the rotation matrix to the rotation vector, so that the rotation increment can be expressed by a three-dimensional variable, so that the parameter expression is minimized, and the error state variable is used to update the filter, which is far away from the singular point and more stable in terms of value, thereby realizing multi-sensor fusion positioning and significantly improving the positioning accuracy and stability of the robot in different underwater motion modes;

[0101] In the underwater stealth mode, the present invention combines the Doppler acoustic Doppler velocimeter DVL and the inertial measurement unit IMU data to provide high-precision underwater positioning; in the underwater crawling mode, the odometer and IMU data are fused to achieve stable positioning.

[0102] By real-time switching and fusing sensor data, this method provides a continuous and reliable navigation solution in complex underwater environments, overcoming the problem of insufficient positioning accuracy of existing technologies in multiple environments and multiple modes. It is widely applicable to ocean exploration, environmental monitoring and other fields, and greatly improves the actual application effect and task execution capability of underwater robots. BRIEF DESCRIPTION OF THE DRAWINGS

[0103] Figure 1 It is a schematic diagram of the flow chart of the combined navigation method of the present invention. DETAILED DESCRIPTION

[0104] The present invention is further described in detail below in conjunction with the accompanying drawings.

[0105] according to Figure 1, the present invention proposes an amphibious robot integrated navigation method based on an error state Kalman filter (ESKF). The initial zero bias, position and attitude parameters of the integrated navigation are determined by initial alignment. The robot uses multiple sensors for high-precision positioning in different modes: in the stealth mode, a Doppler velocimeter (DVL) is used to obtain speed information, and combined with the inertial measurement unit (IMU) data; in the crawling mode, an odometer (OD) is used to obtain motion information, and stable positioning is achieved by combining IMU data. By switching and fusing sensor data in real time, the error state is corrected, and the nominal state is updated to ensure the continuous and reliable navigation of the robot in various complex underwater environments, and improve positioning accuracy and stability.

[0106] A combined navigation method for an underwater robot capable of diving and crawling, the method comprising the following steps:

[0107] Step (1): Initial alignment of the navigation system: The conversion relationship between the carrier coordinate system and the navigation coordinate system is determined by using the accelerometer and gyroscope to obtain the initial attitude matrix.

[0108] Step (2): Establishment of state transfer equation: The rotation matrix is ​​the key state variable in integrated navigation. Based on the Lie group and Lie algebra method, the rotation matrix is ​​mapped to the rotation vector and the IMU track is calculated, thereby realizing the establishment of the state transfer equation.

[0109] Step (3): Establishment of observation equations: Establish observation equations with DVL as the observation source and observation equations with odometer as the observation source respectively.

[0110] Step (4) data fusion: The state transfer equation established in step (2) is used to perform the prediction process of the ESKF and the observation equation established in step (3) is used to perform the update process of the ESKF, thereby achieving fusion positioning under different working conditions (crawling / sneaking).

[0111] The initial alignment step of the navigation system includes the following steps:

[0112] Step (11) At the initial moment of alignment, define the carrier inertial system b0 and the navigation system n0. The b0 system coincides with the carrier coordinate system (b system) at the start of alignment, and then does not rotate relative to the inertial space. The n0 system coincides with the navigation coordinate system (n system) at the start of alignment, and then does not rotate relative to the inertial space. Initial attitude matrix The overall breakdown is as follows:

[0113]

[0114] Step (12) for the submatrix have:

[0115]

[0116] In the formula, It is the angular velocity vector of the Earth's rotation, and can also represent the fixed-axis rotation of the n system relative to the n0 system.

[0117]

[0118] ω ie represents the angular velocity of the Earth's rotation, and L represents the local latitude.

[0119] Submatrix The specific calculation method within the time period t is as follows:

[0120]

[0121] Where L represents the local latitude. For the submatrix have:

[0122]

[0123] In the above formula, Represents gyroscope measurement value, attitude matrix The initial value of is set to the identity matrix I.

[0124] Pair Matrix Solve it and get from formula (5) back, The projection under the b0 system is:

[0125]

[0126] From formula (4), we can get Then, we can get the projection of the gravity vector in the n0 system.

[0127]

[0128] Integrate equations (6) and (7) and record:

[0129]

[0130] In the above formula, i = 1, 2, usually t2 = 2t1, t2 is the end time of initial alignment. At the same time, the relationship between the inertial guidance specific force and the gravity vector is given:

[0131]

[0132] The superscript ~ indicates the actual measured value. It represents the accelerometer measurement error and linear acceleration interference in the b0 system.

[0133] Integrate equation (9) and substitute equation (8) into it to obtain:

[0134]

[0135] neglect The interference of , and the double vector attitude determination method, can be obtained The final result:

[0136]

[0137] Furthermore, the step of establishing the state transfer equation includes the following steps:

[0138] Step (21) The mapping relationship from rotation matrix (Lie group) to rotation vector (Lie algebra):

[0139] According to BCH approximation:

[0140]

[0141] On the contrary:

[0142] exp((Φ+△Φ) ∧ )=exp((J l △Φ) ∧ )exp(Φ ∧ )=exp(Φ ∧ )exp((J r △Φ) ∧ ) (13)

[0143] in:

[0144]

[0145]

[0146] J r (φ)=J l (-φ) (16)

[0147] Step (22) sets the state variables and uses the manifold method to find the Jacobian matrix F of the error state variables:

[0148] Set the state variable x = [p, v, R, b g ,b a ,g] T , where p is translation, v is velocity, R is rotation matrix, b g ,b ais the gyroscope and the table zero bias, g is gravity. Translation, velocity, and rotation matrix are necessary information in navigation and need not be elaborated. Zero bias has a great influence on sensor error and will accumulate and need to be compensated. Gravity acceleration changes in a dynamic environment and needs to be compensated.

[0149] Discrete-time equation of motion for the error state variable:

[0150]

[0151] The corresponding Jacobian matrix F is:

[0152]

[0153] Furthermore, the step of establishing the observation equation includes the following steps:

[0154] Step (31) The observation equation with DVL as the observation source is established:

[0155] The velocity observation of DVL is regarded as the velocity observation of the x, y, and z axes under the carrier system. Then, the velocity observation in the vector form under the navigation coordinate system can be written as:

[0156] v dvl =[v x ,v y ,v z ] (19)

[0157] where v x ,v y ,v z Represents the three-axis velocity components in the navigation coordinate system. The corresponding observation model is:

[0158] z dvl =v+δv+ω dvl (20)

[0159] Where v represents the nominal state variable, δv represents the error state variable, ω dvl represents the DVL observation noise. Obviously, its Jacobian matrix for the error variable δx is Step (32) The observation equation is established with the odometer as the observation source:

[0160] Taking the upper left coordinate system, the wheel speed observation in vector form can be recorded as:

[0161] v od =[v od ,0,0] T (twenty one)

[0162] The corresponding observation model is:

[0163] v od=R T v (22)

[0164] Among them, R,v is the current state of the carrier. Therefore, the odometer can be regarded as a T However, in practice, the odometer does not physically measure R, so the more common method is to use v according to the current estimate of R. od Switch to the navigation coordinate system and directly regard it as the observation of v:

[0165] R od =v (23)

[0166] The corresponding observation model is:

[0167] z od =v+δv+ω od (twenty four)

[0168] Where v represents the nominal state variable, δv represents the error state variable, ω od represents the odometry observation noise.

[0169] Obviously, its Jacobian matrix for the error variable δx is

[0170] Furthermore, the data fusion step includes the following steps:

[0171] Step (41) establishes the prediction equation of the state variables, including the prediction of the nominal state variables and the error state

[0172] Prediction of variables:

[0173]

[0174]

[0175] The prediction of the nominal state only includes the prediction of the state variables, and the prediction of the error state includes the error state variables and the corresponding covariance prediction.

[0176] Step (42): Perform the ESKF update process according to the observation equations corresponding to different working conditions (crawling / sneaking):

[0177] While sneaking:

[0178]

[0179] While crawling:

[0180]

[0181] Step (43) combines the error state with the nominal state to obtain the final fused positioning data:

[0182] After the prediction and update process, the error state estimate is corrected. Next, the error state is simply classified into the nominal state and the ESKF is reset. The final combined navigation positioning information can be expressed as:

[0183]

Claims

1. A stealth and crawling combined navigation method for an amphibious robot, characterized in that: The steps include: Step (1): Initial alignment of the navigation system: Determine the conversion relationship between the carrier inertial coordinate system and the navigation coordinate system through the accelerometer and gyroscope to obtain the initial attitude matrix; Step (2): Establishing the state transfer equation: Mapping the rotation matrix to the rotation vector based on the Lie group Lie algebra method and performing dead reckoning on the inertial measurement unit IMU to establish the state transfer equation; Step (3): Establish observation equations: Establish observation equations with the Doppler velocity meter DVL as the observation source and observation equations with the odometer OD as the observation source respectively; Step (4): Data fusion: Use the state transfer equation established in step (2) to perform the prediction process of the error state Kalman filter ESKF, and use the observation equation established in step (3) to perform the update process of the error state Kalman filter ESKF to achieve fusion positioning under different working conditions.

2. The method according to claim 1, characterized in that Step (1) comprises the following steps: Step (11): Decomposing the initial posture matrix to be solved; Step (12): gradually solve each sub-matrix of the decomposed initial posture matrix to obtain the initial posture matrix and complete the initial alignment.

3. The method according to claim 2, characterized in that Step (11) is specifically: At the initial moment of alignment, the carrier inertial coordinate system b0 and the navigation coordinate system n0 are defined. The carrier inertial coordinate system b0 coincides with the carrier inertial coordinate system b at the start of alignment, and then there is no rotation relative to the inertial space. The n0 system coincides with the navigation coordinate system n at the start of alignment, and then there is no rotation relative to the inertial space. The initial attitude matrix The overall breakdown is as follows:

4. The method according to claim 3, characterized in that Step (12) is specifically as follows: For the submatrix In the formula, is the angular velocity vector of the Earth's rotation, which means the fixed-axis rotation of the n system relative to the n0 system ω ie represents the magnitude of the Earth's rotation angular velocity, and L represents the local latitude; Submatrix The calculation method in the time period t is as follows: Where L represents the local latitude; for In the formula, Represents gyroscope measurement value, attitude matrix The initial value of is set to the unit matrix I; Pair Matrix Solve it and get from formula (5) back, The projection under the b0 system is From formula (4), we can get Then, we get the projection of the gravity vector in the n0 system. Integrate equations (6) and (7) and record: In the formula, i = 1, 2, t2 = 2t1, t2 is the end time of initial alignment, and the relationship between the inertial navigation specific force and the gravity vector is given as: The superscript ~ indicates the actual measured value. It represents the accelerometer measurement error and linear acceleration disturbance in the b0 system; Integrate equation (9) and substitute equation (8) into it to obtain: neglect The interference of the double vector method is used to determine the attitude, and we get The final result 5. The method according to claim 4, characterized in that Step (2) specifically includes the following steps: Step (21): Mapping relationship from rotation matrix to rotation vector: According to BCH approximation: On the contrary: exp((Φ+△Φ) ∧ )=exp((J l △Φ) ∧ )exp(Φ ∧ )=exp(Φ ∧ )exp((J r △Φ) ∧ ) (13) in: J r (φ)=J l (-f) (16) Step (22): Set the state variables and use the manifold method to find the Jacobian matrix F of the error state variables: Set the state variable x = [p, v, R, b g ,b a ,g] T , where p is translation, v is velocity, R is rotation matrix, b g ,b a is the gyroscope and the table zero bias, g is the gravity; Discrete-time equation of motion for the error state variable: The corresponding Jacobian matrix F is:

6. The method according to claim 5, characterized in that: Step (3) specifically includes the following steps: Step (31): Establish the observation equation with DVL as the observation source: The velocity observation of DVL is regarded as the velocity observation of the x, y, and z axes under the carrier system. Then, the velocity observation in the vector form under the navigation coordinate system is written as: v dvl =[v x ,v y ,v z ] (19) where v x ,v y ,v z Represents the three-axis velocity components in the navigation coordinate system; The corresponding observation model is: z dvl =v+δv+ω dvl (20) Where v represents the nominal state variable, δv represents the error state variable, ω dvl represents the DVL observation noise; Observation Model Z dvl The Jacobian matrix for the error variable δx Step (32): Establish the observation equation with the odometer as the observation source: Taking the upper left coordinate system, the wheel speed observation in vector form is recorded as: v od =[v od ,0,0] T (21) The corresponding observation model is: v od =R T v (22) Among them, R,v is the current state of the vehicle. According to the current estimated R, v od Switch to the navigation coordinate system and directly regard it as the observation of v: Rv od =v(23) The corresponding observation model is: z od =v+δv+ω od (24) Where v represents the nominal state variable, δv represents the error state variable, ω od represents the odometer observation noise; observation model Z od The Jacobian matrix for the error variable δx 7. The method according to claim 6, characterized in that: Step (4) comprises the following steps: Step (41): Establish the prediction equation of the state variables, including the prediction of the nominal state variables and the prediction of the error state variables: Step (42): Update the ESKF according to the observation equation corresponding to the crawling or sneaking condition: While sneaking: While crawling: Step (43) combines the error state with the nominal state to obtain the final fused positioning data: After prediction and updating, the error state estimate is corrected, the error state is classified as the nominal state, and then the ESKF is reset. The final integrated navigation positioning information is expressed as:

8. A stealth and crawling combined navigation system for an amphibious robot, characterized in that: Navigation is performed using the navigation method described in any one of claims 1 to 7.

9. The system according to claim 8, characterized in that Including Doppler acoustic Doppler velocity meter DVL, inertial measurement unit IMU, odometer OD and error state Kalman filter; The error state Kalman filter has an observation model Z with the Doppler acoustic Doppler velocity meter DVL as the observation source dvl And the observation model Z with odometer OD as the observation source od ; In stealth mode, the Doppler acoustic Doppler velocimeter DVL is used to obtain velocity information, combined with the inertial measurement unit IMU data, and the error state Kalman filter uses the observation model Z dvl ; In the crawling mode, the odometer OD is used to obtain motion information, combined with the IMU data, and the error state Kalman filter uses the observation model Z od .

Citation Information

Cited By

  • Multi-sensor fusion underwater robot positioning method and system

    CN119178430A

  • A multi-sensor fusion underwater robot positioning method and system

    CN119178430B

  • Intelligent amphibious ship navigation positioning method and system and storage medium

    CN119845287A

  • Internet of Things terminal multi-source adaptive fusion positioning system and method

    CN121207167A

  • Internet of things terminal multi-source adaptive fusion positioning system and method

    CN121207167B