CN-122015810-A - UUV inertial navigation-electromagnetic fusion positioning method and system based on magnetic gradient tensor

CN122015810ACN 122015810 ACN122015810 ACN 122015810ACN-122015810-A

Abstract

The application discloses a UUV inertial navigation-electromagnetic fusion positioning method and system based on magnetic gradient tensor, which relate to the field of underwater navigation positioning and comprise the steps of obtaining electromagnetic beacon magnetic field signals, UUV angular speed and linear acceleration signals, preprocessing, calculating independent components of the magnetic gradient tensor by adopting a space difference method, resolving the motion state of an inertial navigation system to obtain position, speed, gesture and IMU errors of the UUV, establishing a tightly coupled fusion model of the magnetic gradient tensor and the motion state of the inertial navigation system based on a magnetic dipole equivalent model, carrying out state estimation on the tightly coupled fusion model by adopting an unscented Kalman filtering algorithm, correcting the state of the inertial navigation system by taking the independent components of the magnetic gradient tensor as observation values, and outputting a fusion positioning result of the UUV.

Inventors

LI HENG
ZHAO LIANGHAO
ZHU REN
LI ZIYUAN
YANG WENTIE
WANG JIANXUN
WANG ZUOSHUAI

Assignees

中南大学

Dates

Publication Date: 20260512
Application Date: 20260408

Claims (10)

1. A UUV inertial navigation-electromagnetic fusion positioning method based on magnetic gradient tensor is characterized by comprising the following steps: S1, acquiring an electromagnetic beacon magnetic field signal, a UUV angular velocity signal and a linear acceleration signal; s2, preprocessing electromagnetic beacon magnetic field signals, UUV angular speed signals and linear acceleration signals, wherein the preprocessing comprises zero offset calibration and filtering processing; s3, calculating independent components of magnetic gradient tensors by a spatial difference method based on the preprocessed electromagnetic beacon magnetic field signals; S4, based on the preprocessed UUV angular velocity and linear acceleration signals, calculating the motion state of the inertial navigation system to obtain the position, velocity, attitude and IMU error of the UUV; s5, based on a magnetic dipole equivalent model, establishing a tightly coupled fusion model of a magnetic gradient tensor and a motion state of an inertial navigation system, wherein the tightly coupled fusion model comprises an inertial navigation system state equation and a magnetic gradient tensor observation equation; And S6, carrying out state estimation on the tightly coupled fusion model by adopting a unscented Kalman filtering algorithm, correcting the state of the inertial navigation system by using the independent component of the magnetic gradient tensor obtained in the step S3 as an observation value, and outputting a fusion positioning result of the UUV.
2. The UUV inertial navigation-electromagnetic fusion positioning method based on the magnetic gradient tensor according to claim 1 is characterized in that electromagnetic beacon magnetic field signals are generated by 3 fixed electromagnetic beacons distributed on the sea floor, each electromagnetic beacon is equivalent to a magnetic dipole perpendicular to the sea floor plane, the electromagnetic beacon magnetic field signals are collected by a miniature magnetic gradient tensor observation system, the miniature magnetic gradient tensor observation system comprises four triaxial magnetometers, and the miniature magnetic gradient tensor observation system is distributed on a UUV carrier in a regular tetrahedral array.
3. The UUV inertial navigation-electromagnetic fusion positioning method based on magnetic gradient tensor according to claim 2, wherein the preprocessing of the electromagnetic beacon magnetic field signal specifically comprises: zero offset calibration is carried out on the acquired electromagnetic beacon magnetic field signals, and fixed zero offset errors of the magnetometer are removed; And carrying out sliding window average filtering on the electromagnetic beacon magnetic field signal subjected to zero offset calibration to remove random noise and pulse interference.
4. A UUV inertial navigation-electromagnetic fusion positioning method based on magnetic gradient tensor according to claim 3, wherein the preprocessing of UUV angular velocity and linear acceleration signals specifically comprises: zero offset calibration is carried out on UUV angular velocity and linear acceleration signals, and fixed zero offset of the sensor is removed; And carrying out low-pass filtering on the UUV angular velocity and the linear acceleration signal after zero offset calibration to remove high-frequency noise.
5. The UUV inertial navigation-electromagnetic fusion positioning method based on magnetic gradient tensor according to claim 4, wherein calculating the independent components of the magnetic gradient tensor by using the spatial difference method specifically comprises: Calculating the space distance between any two magnetometers under a north-east coordinate system according to the pre-calibrated relative coordinates of the magnetometers; Calculating the magnetic field intensity difference value of adjacent magnetometers in all directions based on the preprocessed electromagnetic beacon magnetic field signals; dividing the magnetic field intensity difference by the corresponding space interval to obtain the space partial derivative of the magnetic field in each direction, and constructing a magnetic gradient tensor matrix; five independent components in the magnetic gradient tensor matrix are extracted as observation parameters of electromagnetic positioning.
6. The UUV inertial navigation-electromagnetic fusion positioning method based on magnetic gradient tensor according to claim 5, wherein the inertial navigation system motion state calculation comprises: describing an evolution rule of the UUV motion state through a nonlinear state equation based on the preprocessed angular velocity and linear acceleration signals; and solving a state equation by adopting a Dragon-Kutta integration method to obtain the position, the speed, the gesture and the IMU error of the UUV.
7. The UUV inertial navigation-electromagnetic fusion positioning method based on magnetic gradient tensor according to claim 6, wherein the state equation of the inertial navigation system describes the evolution law of the UUV motion state, and is established based on the principle of rigid body kinematics, expressed as: ; Wherein, the For the inertial navigation system state vector at time k, For the preprocessed angular velocity and linear acceleration signals at time k-1, As a function of the non-linear kinematic mapping, Zero mean gaussian process noise; the magnetic gradient tensor observation equation is used for establishing a direct mapping relation between the state of the inertial navigation system and the magnetic gradient tensor, and is expressed as follows: ; Wherein, the For the measured values of 5 independent components of the magnetic gradient tensor at time k, To observe the mapping function, extract Middle position And a gesture The theoretical gradient component is calculated in combination with the magnetic dipole model, Is zero-mean Gaussian observed noise.
8. The UUV inertial navigation-electromagnetic fusion positioning method based on magnetic gradient tensor according to claim 7, wherein the state estimation of the tightly coupled fusion model is performed by adopting a unscented kalman filter algorithm, specifically comprising: Initializing an initial state estimation value and a covariance matrix; Generating Sigma points based on the state estimation value and the covariance matrix of the last moment; Substituting the generated Sigma points into an inertial navigation system state equation to perform prediction step recurrence to obtain a state prediction value and a covariance prediction matrix; substituting Sigma points obtained by recursion of the prediction step into a magnetic gradient tensor observation equation, and calculating an observation prediction mean value, an observation covariance matrix and a state-observation cross covariance matrix; calculating Kalman gain, and correcting the state predicted value by using the independent component of the magnetic gradient tensor obtained in the step S3 as an observed value to obtain a final state estimated value; and extracting the three-dimensional position and speed of the UUV from the final state estimation value to be used as a fusion positioning result.
9. The UUV inertial navigation-electromagnetic fusion positioning method based on magnetic gradient tensor according to claim 8, wherein step S6 further comprises: And taking the state estimation value and the covariance matrix obtained at the current moment as the filtering input at the next moment to realize continuous positioning.
10. A UUV inertial navigation-electromagnetic fusion positioning system based on magnetic gradient tensor, comprising: The submarine electromagnetic beacon module is used for providing an absolute magnetic field reference and comprises a plurality of electromagnetic beacons fixedly distributed on the seabed in a UUV operation area, wherein each beacon is equivalent to a magnetic dipole; The UUV vehicle-mounted sensing module comprises a miniature magnetic gradient tensor observation unit and an inertia measurement unit, wherein the miniature magnetic gradient tensor observation unit is a tetrahedron array formed by a plurality of triaxial magnetometers and is used for collecting electromagnetic beacon magnetic field signals generated by electromagnetic beacons, and the inertia measurement unit is used for collecting angular speed and linear acceleration signals of the UUV; The fusion positioning calculation module comprises a signal preprocessing unit, a magnetic gradient tensor calculation unit, an INS (inertial navigation System) calculation unit, a fusion modeling unit, an UKF (unscented Kalman Filter) unit and a positioning result unit, and is used for receiving and processing electromagnetic beacon magnetic field signals, UUV (unmanned underwater vehicle) angular velocity and linear acceleration signals, executing the UUV inertial navigation-electromagnetic fusion positioning method based on the magnetic gradient tensor according to any one of claims 1 to 9 and outputting a UUV fusion positioning result.

Description

UUV inertial navigation-electromagnetic fusion positioning method and system based on magnetic gradient tensor Technical Field The application relates to the technical field of underwater navigation positioning, in particular to a UUV inertial navigation-electromagnetic fusion positioning method and system based on magnetic gradient tensor. Background The unmanned underwater vehicle (Unmanned Underwater Vehicle, UUV) is core equipment for tasks such as marine resource exploration, submarine engineering inspection, underwater security search and rescue and the like, and high-precision and long-time autonomous positioning is a key for realizing underwater autonomous operation. The satellite navigation signal in the underwater environment is completely invalid, the inertial navigation system (Inertial Navigation System, INS) can realize high-frequency and continuous pose output, so that the inertial navigation system becomes a core basic means for underwater positioning of the UUV, but the INS is recursively calculated based on the measured value of the inertial measurement unit (Inertial Measurement Unit, IMU), the problem that inherent errors are accumulated along with time exists, and the single INS navigation cannot meet the positioning precision requirement of the UUV for long-term operation. In order to restrain the INS from accumulating drift errors, multi-sensor fusion positioning becomes a mainstream technical direction, wherein the INS is fused with electromagnetic positioning with absolute position reference capability, and the multi-sensor fusion positioning is an ideal scheme for adapting UUV underwater operation. The underwater electromagnetic positioning is a preferable absolute positioning reference for INS fusion positioning by virtue of the advantages of stable electromagnetic signal propagation, no influence of sea water turbidity temperature change, low beacon arrangement cost, strong concealment and the like. The existing underwater electromagnetic positioning technology mostly uses single magnetic field intensity as an observation parameter, has the inherent defects of weak anti-interference, low utilization rate of observation information, short effective positioning distance and the like, can not inhibit UUV carrier uniform magnetic interference and underwater environment uniform magnetic disturbance from a physical layer, has large observation signal deviation, and is difficult to be used as a reliable absolute positioning reference. Traditional UUV inertial navigation-electromagnetic fusion positioning is performed by loose coupling fusion based on electromagnetic positioning results of magnetic field intensity, a direct coupling model of inertial navigation motion state and electromagnetic magnetic field characteristics is not established, observation errors of electromagnetic positioning are easy to amplify, and fusion positioning accuracy is limited. In the prior art, although there are methods for exploring magnetic field information for positioning or tracking, such as magnetic field inversion positioning, magnetic target positioning, weighted extended Kalman filtering fusion and other schemes, and magnetic beacon and inertial sensor fusion positioning systems proposed in international patents, the core problems of insufficient electromagnetic positioning self-interference resistance, error amplification of fusion mode, limited effective operation range and the like are not solved in the prior art, and the performance improvement of UUV underwater high-precision long-time positioning is restricted. Disclosure of Invention In order to solve the problems, the application provides a UUV inertial navigation-electromagnetic fusion positioning method and system based on magnetic gradient tensor, which realize the omnibearing improvement of the inertial navigation-electromagnetic fusion positioning performance by optimizing the self performance of electromagnetic positioning and establishing a high-precision tight coupling fusion model with INS. In order to achieve the above purpose, the application provides a UUV inertial navigation-electromagnetic fusion positioning method based on magnetic gradient tensor, which comprises the following steps: S1, acquiring an electromagnetic beacon magnetic field signal, a UUV angular velocity signal and a linear acceleration signal; s2, preprocessing electromagnetic beacon magnetic field signals, UUV angular speed signals and linear acceleration signals, wherein the preprocessing comprises zero offset calibration and filtering processing; s3, calculating independent components of magnetic gradient tensors by a spatial difference method based on the preprocessed electromagnetic beacon magnetic field signals; S4, based on the preprocessed UUV angular velocity and linear acceleration signals, calculating the motion state of the inertial navigation system to obtain the position, velocity, attitude and IMU error of the UUV; s5, based on a magnetic dipole