An underwater AUV positioning system and method of noise distribution modeling

Abstract

The application relates to an underwater AUV positioning system and method for noise distribution modeling, and comprises the following steps: performing slant range and horizontal angle measurement by using an underwater acoustic base station arranged on a water surface ship and an underwater beacon arranged on an underwater AUV; performing motion parameter measurement by using an inertial measurement unit arranged on the underwater AUV; considering the influence of unmodeled noise on positioning accuracy, performing approximate modeling on non-Gaussian noise of the slant range, the horizontal angle and the motion parameter; combining the depth of the underwater AUV to establish a nonlinear state equation and a nonlinear measurement equation capable of representing the position of the underwater AUV; and performing state updating of the underwater AUV based on an information matrix and outputting a positioning result. The application considers the influence of non-Gaussian noise on the positioning accuracy of the underwater AUV, performs noise modeling preprocessing, and then uses information filtering to perform underwater AUV positioning estimation, so that technical support can be provided for applications including underwater multi-machine cooperation.

Description

Technical Field

[0001] This invention relates to the field of underwater vehicle positioning and navigation technology, and in particular to an underwater AUV positioning system and method based on noise distribution modeling. Background Technology

[0002] As marine resource development extends into the deep sea, underwater robots, as crucial vehicles for exploring the ocean, acquiring marine resources, and maintaining marine ecological security, are indispensable in fields such as seabed mineral exploration, deep-sea environmental monitoring, underwater reconnaissance operations, and marine emergency rescue. As a key foundational technology of the integrated marine positioning, navigation, and timing (PNT) system, the positioning performance of underwater robots determines whether a target can move to the designated operational area according to mission requirements. However, considering the complex and variable marine environment, the instability of underwater sensing measurements, and the severe attenuation of electromagnetic waves by the water medium rendering satellite navigation unsuitable, the introduction of new positioning algorithms and data processing methods to improve the positioning performance of underwater robots has become a prerequisite for ensuring the effectiveness of various marine engineering applications.

[0003] As the only means of long-distance information transmission in the ocean, underwater robot positioning can be based on underwater acoustic communication. By measuring the time or phase of sound pulses propagating along different paths, the three-dimensional position of the underwater robot can be calculated. However, due to limitations such as the instability of underwater acoustic communication links, the curvature of underwater acoustic velocity profiles, and large transmission delays, target positioning based on underwater acoustic communication is prone to significant positioning errors in some areas. To address this, inertial navigation or geophysical navigation can be combined with underwater acoustic positioning. By fusing multi-source heterogeneous information, the problems of single-sensor measurement can be suppressed, enhancing the positioning performance of underwater robots. Different sensing measurement methods have different advantages and disadvantages. Inertial navigation, as a method that does not rely on external signals, calculates the target's velocity, position, and attitude through three-axis angular acceleration and acceleration integral. However, long-term accumulated errors require correction by external sensors. Geophysical navigation includes terrain matching, geomagnetic matching, and gravity matching. Its principle is to compare the measured local seabed topography, geomagnetic data, or gravity field with a pre-built reference database and find the optimal matching position through relevant algorithms. However, its positioning accuracy is highly dependent on the resolution, coverage, and timeliness of the prior database.

[0004] Regardless of the positioning technology used, underwater robot positioning is used in marine science and engineering practices such as deep-sea mining, submarine cable laying, shipwreck salvage, and ecological monitoring. With the penetration of deep learning and adaptive filtering algorithms into the field of underwater robot positioning, underwater robot combined positioning can achieve higher robustness, providing technical support for underwater multi-machine collaboration and even unmanned autonomous movement. Summary of the Invention

[0005] To address the shortcomings of existing production technologies, the applicant provides an underwater AUV positioning system and method based on noise distribution modeling, which effectively solves the problem of unstable positioning accuracy of existing underwater AUVs and improves the reliability of underwater AUV positioning.

[0006] The technical solution adopted in this invention is as follows: An underwater AUV positioning system based on noise distribution modeling includes a surface vessel, an underwater AUV, an underwater acoustic base station, an underwater beacon, an inertial measurement unit (IMU), a depth sensor, and a positioning calculation module. The slant range and horizontal angle of the underwater AUV relative to the surface vessel are measured by the IMU installed on the surface vessel and the underwater beacon installed on the underwater AUV. The motion parameters of the underwater AUV are collected by the IMU installed on the fuselage. Considering the impact of unmodeled noise on positioning accuracy, non-Gaussian noise is processed to construct a mapping between the underwater AUV's position and the slant range, horizontal angle, and motion parameters. Combining the underwater AUV's depth measured by the depth sensor, a nonlinear state equation and a nonlinear measurement equation characterizing the underwater AUV's position are established. The underwater AUV's position is then obtained by the positioning calculation module installed in the computer inside the underwater AUV.

[0007] A noise distribution modeling method for underwater AUV localization includes the following steps: Step 1: Deploy a surface vessel and an underwater AUV. The surface vessel is equipped with an underwater acoustic base station, and the underwater AUV is equipped with an underwater beacon, an inertial measurement unit (IMU), and a depth sensor. The slant distance and horizontal angle of the underwater AUV relative to the surface vessel are measured through underwater acoustic communication between the base station and the IMU. The three-axis velocity and three-axis position of the underwater AUV are obtained through the IMU. The depth of the underwater AUV is measured through the depth sensor. Step 2: Establish the nonlinear state equation and nonlinear measurement equation for the underwater AUV; the nonlinear state equation uses the three-axis velocity and three-axis position of the underwater AUV as state variables, and the three-axis acceleration as input to model the state noise distribution; the nonlinear measurement equation uses the three-dimensional position of the underwater AUV as the estimation variable, and the slant range, horizontal angle and depth as input to model the measurement noise distribution. Step 3: Based on the similarity of measurement values ​​within the neighborhood space during continuous underwater AUV movement, obtain the values ​​within the neighborhood space. Set of measured values, calculate The estimated positions of multiple underwater AUVs are determined by weighting coefficients based on the covariance of the measurement noise of each group, and the estimated positions of multiple groups in the neighborhood space are weighted and fused to update the nonlinear measurement equation. Step 4: Perform a first-order Taylor expansion on the updated nonlinear measurement equation at the state vector prediction estimate to obtain an approximately linearized measurement equation; calculate the information matrix increment and information vector increment, update the state vector and covariance matrix of the underwater AUV based on the information filtering algorithm, and output the position of the underwater AUV.

[0008] Its further technical solution lies in: Underwater AUVs equipped with inertial measurement units The state at time is The three-axis velocity and three-axis position components are respectively and Underwater AUVs in Nonlinear state equation at time 1 for:

[0009] In the formula: Here is the state transition matrix of an underwater AUV. ; This is the input matrix for an underwater AUV. ; It is a triaxial acceleration; Non-Gaussian state noise is considered. Given that unmodeled non-Gaussian state noise reduces positioning accuracy, non-Gaussian state noise is addressed... The distribution model is as follows:

[0010] In the formula: This represents the mean of the state noise; is the scaling matrix of the state noise; For the degree of freedom parameter; introduce random variables. ,exist time Represented as containing random variables The distribution is as follows:

[0011] In the formula: The mean is The sum and variance are Gaussian distribution; Indicates compliance with parameters and The gamma distribution.

[0012] The reference coordinates of a surface vessel equipped with an underwater acoustic base station are as follows: Underwater AUVs equipped with underwater beacons are... The coordinates to be estimated at time 1 are Underwater beacons and underwater acoustic base stations in The slant range and horizontal angle are obtained through underwater acoustic communication at various times. and The underwater AUV equipped with a depth sensor has a depth of [missing information]. Establish a nonlinear measurement equation that can characterize the position of an underwater AUV. for

[0013] In the formula: The horizontal angle between the underwater acoustic base station and the underwater beacon; The slant distance between the underwater acoustic base station and the underwater beacon; The depth of the underwater AUV; Non-Gaussian measurement noise; considering that unmodeled non-Gaussian measurement noise reduces positioning accuracy, non-Gaussian measurement noise is... The distribution model is as follows:

[0014] In the formula: To measure the mean of the noise; The scaling matrix for measuring noise; For the degree of freedom parameter; introduce random variables. ,exist time Represented as containing random variables The distribution is as follows:

[0015] In the formula: The mean is The sum and variance are Gaussian distribution; Indicates compliance with parameters and The gamma distribution.

[0016] Because of the continuous movement of underwater AUVs, similar measurement values ​​will be obtained in the neighborhood space. The measured value is expressed as ,in Indicates that there is in the neighborhood space Group of measurements; based on the first set of measurements Slope distance measured in the second measurement Horizontal angle ,depth It can be solved Estimated location of the group of underwater AUVs for:

[0017] The measurement equation based on neighborhood space characterizing the location of underwater AUVs can be updated as follows:

[0018] In the formula: These are the weighting coefficients. For the first Covariance of the second measurement noise; updated nonlinear measurement equation and measurement noise covariance Substitute into the location solution module; for the equation In the state vector prediction estimate A first-order Taylor expansion yields an approximately linearized measurement equation:

[0019] In the formula: To approximate linearize the measurement equation Regarding the estimated value of the state vector prediction The Jacobian matrix. This will approximate the linearization of the measurement equation. Substituting into the nonlinear measurement equation We can obtain:

[0020] The information matrix increment and information vector increment used for underwater AUV positioning are as follows:

[0021] In the formula: The equivalent noise covariance matrix; the state vector of the underwater AUV. and corresponding covariance matrix The updates are as follows:

[0022] In the formula: , The above steps will output the estimated position of the underwater AUV. .

[0023] The beneficial effects of this invention are as follows: I. This invention incorporates state noise and measurement noise into the distributed modeling, thereby achieving preprocessing of non-Gaussian noise.

[0024] Existing underwater AUV positioning methods are generally based on the Gaussian noise assumption. However, in actual underwater environments, due to factors such as marine ambient noise, underwater acoustic signal fading, multipath effects, and sensor malfunctions, state noise and measurement noise often exhibit significant non-Gaussian characteristics. When the actual noise deviates from the Gaussian distribution, the estimation accuracy of traditional Kalman filtering and its variants will decrease significantly, and may even lead to filter divergence.

[0025] Some studies have attempted to address non-Gaussian measurement noise, but most only model the measurement noise as non-Gaussian, neglecting the fact that state noise in the AUV motion model also exhibits non-Gaussian properties. For example, CN107677272B uses an unscented information filter to handle nonlinear problems, but implicitly assumes Gaussianism; while the cooperative navigation method based on robust information filtering considers anomalous measurement noise, it indirectly suppresses its influence through robustness, without modeling the statistical characteristics of state noise and measurement noise.

[0026] This invention addresses this technical deficiency by accurately modeling the distribution of state noise and measurement noise: transforming non-Gaussian noise into an approximate Gaussian combination, enabling effective output of the prediction and update steps of information filtering. Regardless of the positioning technology used, underwater robot positioning is crucial in marine science and engineering practices such as deep-sea mining, submarine cable laying, shipwreck salvage, and ecological monitoring, impacting the operational efficiency and task coordination of marine equipment.

[0027] Second, this invention proposes a multi-sensor measurement fusion method based on an information matrix, which overcomes the limitation of traditional methods that only utilize a single measurement.

[0028] Existing underwater AUV positioning methods typically update measurements using only the single measurement at the current moment, ignoring the similarity of continuous motion measurements of underwater AUVs: that is, the position of the AUV changes little at adjacent moments, and multiple sets of measurements in the neighborhood space are similar in amplitude.

[0029] This invention utilizes the correlation between multiple sets of measurements within a neighborhood space: based on the correlation between multiple sets of measurements within a neighborhood space. The estimated location of the underwater AUV was calculated from the sets of measurements. Weighting coefficients were constructed based on the noise covariance matrix of each measurement and then weighted and fused to achieve the fusion of multiple sets of measurements in the neighborhood space.

[0030] The essential difference between this technique and existing methods is that existing methods rely on a single measurement at the current moment, while this invention utilizes multiple measurements within the neighborhood space and effectively suppresses the influence of a single measurement on the positioning result through weighted fusion.

[0031] Third, the three core technical features of this invention are not simply superimposed, but rather progressively layered: (1) Distributed modeling ensures the accuracy of state estimation under non-Gaussian noise environment, providing a reliable basis for subsequent filtering estimation; (2) Neighborhood space fusion utilizes the measurement similarity under continuous underwater AUV movement and suppresses the influence of outliers in non-Gaussian noise through weighted fusion; (3) Information filtering is updated in the form of information matrix and information vector, and the equivalent measurement noise covariance after neighborhood space fusion is included in the update equation, which has high computational efficiency.

[0032] Fourth, the positioning accuracy of the present invention is significantly improved, and the embodiments provide strong support for this.

[0033] To verify the actual effectiveness of the present invention, a comparative experiment was conducted under the same conditions. Figure 3 Location trajectory map and Figure 4 The positioning error diagram illustrates the performance comparison between the present invention and the existing extended Kalman filter (EKF) method.

[0034] Extended Kalman Filter (EKF): The average positioning error is 0.50m. Since EKF is based on the Gaussian noise assumption and all observations use fixed weights, its positioning accuracy is significantly affected when the actual noise exhibits non-Gaussian characteristics.

[0035] The method of this invention has an average positioning error of 0.37m, and the positioning accuracy is improved by about 26% compared with EKF, thus ensuring that underwater AUVs have better positioning accuracy.

[0036] Fifth, this invention does not require additional hardware costs and has good engineering applicability and promotional value.

[0037] All steps of this invention can be deployed in the positioning and calculation module of the computer inside the underwater AUV cabin. Compared with improving positioning accuracy by adding high-precision sensors, this invention achieves improved positioning performance without increasing system hardware, and has good economic and engineering applicability. It is suitable for various underwater AUV platforms equipped with underwater acoustic communication units, inertial measurement units, and depth sensors.

[0038] In summary, the present invention outperforms existing technologies in non-Gaussian noise modeling, neighborhood spatial correlation measurement, underwater target filtering and localization, and improved localization accuracy. Attached Figure Description

[0039] Figure 1 This is a schematic diagram of the positioning system of the present invention.

[0040] Figure 2 This is a flowchart of the positioning method of the present invention.

[0041] Figure 3 This is a positioning trajectory diagram of the underwater AUV of the present invention.

[0042] Figure 4 This is a diagram showing the positioning error of the underwater AUV of this invention. Detailed Implementation

[0043] The specific embodiments of the present invention will now be described with reference to the accompanying drawings.

[0044] like Figures 1-4 As shown, the underwater AUV positioning system with noise distribution modeling in this embodiment includes a surface vessel, an underwater AUV, an underwater acoustic base station, an underwater beacon, an inertial measurement unit (IMU), a depth sensor, and a positioning calculation module. The slant range and horizontal angle of the underwater AUV relative to the surface vessel are measured by the IMU installed on the surface vessel and the underwater beacon installed on the underwater AUV. The motion parameters of the underwater AUV are collected by the IMU installed on the fuselage. Considering the impact of unmodeled noise on positioning accuracy, non-Gaussian noise is processed to construct a mapping between the underwater AUV's position and the slant range, horizontal angle, and motion parameters. Combining the underwater AUV's depth measured by the depth sensor, a nonlinear state equation and a nonlinear measurement equation that characterize the underwater AUV's position are established. The underwater AUV's position is obtained by the positioning calculation module installed in the computer inside the underwater AUV.

[0045] The underwater AUV positioning method based on noise distribution modeling in this embodiment includes the following steps: Step 1: Deploy surface vessels and underwater AUVs. The surface vessel is equipped with an underwater acoustic base station, and the underwater AUV is equipped with an underwater beacon, an inertial measurement unit (IMU), and a depth sensor. The slant range and horizontal angle of the underwater AUV relative to the surface vessel are measured through underwater acoustic communication between the underwater acoustic base station and the underwater beacon. The three-axis velocity and three-axis position of the underwater AUV are obtained through the IMU. The depth of the underwater AUV is measured through the depth sensor. Step 2: Establish the nonlinear state equation and nonlinear measurement equation for the underwater AUV; the nonlinear state equation uses the three-axis velocity and three-axis position of the underwater AUV as state variables, and the three-axis acceleration as input to model the state noise distribution; the nonlinear measurement equation uses the three-dimensional position of the underwater AUV as the estimation variable, and the slant range, horizontal angle and depth as input to model the measurement noise distribution. Step 3: Based on the characteristic that the amplitude of measured values ​​in the neighborhood space is similar during continuous underwater AUV movement, obtain the values ​​within the neighborhood space. The measured values ​​were calculated separately. The estimated location of the underwater AUVs in the group; the weighting coefficients are determined based on the covariance of the measurement noise of each group, and the results are applied to the neighborhood space. The estimated locations are weighted and fused together to update the nonlinear measurement equations; Step 4: Perform a first-order Taylor expansion on the updated nonlinear measurement equation at the state vector prediction estimate to obtain an approximately linearized measurement equation; calculate the information matrix increment and information vector increment, update the state vector and covariance matrix of the underwater AUV based on the information filtering algorithm, and output the position of the underwater AUV.

[0046] Underwater AUVs equipped with inertial measurement units The state at time is The three-axis velocity and three-axis position components are respectively and Underwater AUVs in Nonlinear state equation at time 1 for:

[0047] In the formula: Here is the state transition matrix of an underwater AUV. ; This is the input matrix for an underwater AUV. ; It is a triaxial acceleration; This is non-Gaussian noise. Considering that unmodeled non-Gaussian noise reduces positioning accuracy, [the following is an explanation of the process for handling non-Gaussian noise]. The distribution model is as follows:

[0048] In the formula: This represents the mean of the state noise; is the scaling matrix of the state noise; For the degrees of freedom parameter. Introduce random variables. ,exist time Represented as containing random variables The distribution is as follows:

[0049] In the formula: The mean is The sum and variance are Gaussian distribution; Indicates compliance with parameters and The gamma distribution.

[0050] The reference coordinates of a surface vessel equipped with an underwater acoustic base station are as follows: Underwater AUVs equipped with underwater beacons are... The coordinates to be estimated at time 1 are Underwater beacons and underwater acoustic base stations in The slant range and horizontal angle are obtained through underwater acoustic communication at various times. and The underwater AUV equipped with a depth sensor has a depth of [missing information]. Establish a nonlinear measurement equation that can characterize the position of an underwater AUV. for:

[0051] In the formula: The horizontal angle between the underwater acoustic base station and the underwater beacon; The slant distance between the underwater acoustic base station and the underwater beacon; The depth of the underwater AUV; This refers to non-Gaussian measurement noise. Considering that unmodeled non-Gaussian measurement noise reduces positioning accuracy, [the following is an explanation of the noise level and its implications]. The distribution model is as follows:

[0052] In the formula: To measure the mean of the noise; The scaling matrix for measuring noise; For the degrees of freedom parameter. Introduce random variables. ,exist time Represented as containing random variables The distribution is

[0053] In the formula: The mean is The sum and variance are Gaussian distribution; Indicates compliance with parameters and The gamma distribution.

[0054] Because of the continuous movement of underwater AUVs, similar measurement values ​​will be obtained in the neighborhood space. The measured value is expressed as ,in Indicates that there is in the neighborhood space Group of measurements. Based on the first set of measurements. Slope distance measured in the second measurement Horizontal angle ,depth It can be solved Estimated location of the group of underwater AUVs for

[0055] The measurement equation based on neighborhood space characterizing the location of underwater AUVs can be updated to:

[0056] In the formula: These are the weighting coefficients. For the first The covariance of the measurement noise. The updated nonlinear measurement equation. and measurement noise covariance Substitute this into the location solution module. For the equation... In the state vector prediction estimate A first-order Taylor expansion yields an approximately linearized measurement equation:

[0057] In the formula: To approximate linearize the measurement equation Regarding the estimated value of the state vector prediction The Jacobian matrix. This will approximate the linearization of the measurement equation. Substituting into the nonlinear measurement equation achievable

[0058] The information matrix increment and information vector increment used for underwater AUV positioning are as follows:

[0059] In the formula: This is the equivalent noise covariance matrix. The state vector of the underwater AUV. and corresponding covariance matrix The updates are as follows:

[0060] In the formula: , The above steps will output the estimated position of the underwater AUV. .

[0061] Figure 3 The image shows the positioning trajectory in an embodiment of the present invention. Both the present invention and the Extended Kalman Array (EKF) can locate trajectories, but the present invention exhibits better positioning performance due to its modeling of non-Gaussian noise. Figure 4 The positioning error in this embodiment of the invention is shown below. Compared to the average positioning error of 0.50 m in the Extended Kalman EKF, the average positioning error in this embodiment of the invention is reduced to 0.37 m. Unlike the Extended Kalman EKF, which is based on the Gaussian noise assumption and uses fixed weights for all observations, this embodiment of the invention introduces non-Gaussian noise modeling to improve the positioning accuracy of underwater AUVs.

[0062] It uses multi-source measurements, including underwater acoustic slant range, horizontal angle, and depth, and takes into account that non-Gaussian noise will affect the positioning performance of underwater AUVs, and performs approximate distribution processing on state noise and measurement noise.

[0063] Based on similar measurements in the neighborhood space, an estimation equation that can characterize the position of an underwater AUV is established. The measurement equation in the neighborhood space is updated and a first-order Taylor expansion is performed to obtain the state estimate of the underwater AUV.

[0064] It uses multi-source measurements to establish nonlinear measurement equations and state equations characterizing the position of the underwater AUV, and obtains the position of the underwater AUV through the positioning calculation module of the computer installed in the underwater AUV cabin.

[0065] The above description is an explanation of the present invention and not a limitation thereof. The scope of the present invention is defined by the claims. Within the scope of protection of the present invention, any form of modification may be made.

Claims

1. An underwater AUV positioning system with noise distribution modeling, characterized in that: The system includes a surface vessel, an underwater AUV, an underwater acoustic base station, an underwater beacon, an inertial measurement unit (IMU), a depth sensor, and a positioning and calculation module. The slant range and horizontal angle of the underwater AUV relative to the surface vessel are measured by the underwater acoustic base station installed on the surface vessel and the underwater beacon installed on the underwater AUV. The motion parameters of the underwater AUV are collected by the IMU installed on the fuselage. Considering the impact of unmodeled noise on positioning accuracy, non-Gaussian noise is processed to construct a mapping between the underwater AUV's position and its slant range, horizontal angle, and motion parameters. Combining the underwater AUV's depth measured by the depth sensor, a nonlinear state equation and a nonlinear measurement equation are established to characterize the underwater AUV's position. The positioning and calculation module installed in the computer inside the underwater AUV's cabin then obtains the underwater AUV's position.

2. A noise distribution modeling method for underwater AUV positioning, characterized in that: The following steps are included: Step 1: Deploy a surface vessel and an underwater AUV. The surface vessel is equipped with an underwater acoustic base station, and the underwater AUV is equipped with an underwater beacon, an inertial measurement unit (IMU), and a depth sensor. The slant distance and horizontal angle of the underwater AUV relative to the surface vessel are measured through underwater acoustic communication between the base station and the IMU. The three-axis velocity and three-axis position of the underwater AUV are obtained through the IMU. The depth of the underwater AUV is measured through the depth sensor. Step 2: Establish the nonlinear state equation and nonlinear measurement equation for the underwater AUV; the nonlinear state equation uses the three-axis velocity and three-axis position of the underwater AUV as state variables, and the three-axis acceleration as input to model the state noise distribution; the nonlinear measurement equation uses the three-dimensional position of the underwater AUV as the estimation variable, and the slant range, horizontal angle and depth as input to model the measurement noise distribution. Step 3: Based on the similarity of measurements within the neighborhood space during continuous underwater AUV movement, obtain the measurements within the neighborhood space. Set of measured values, calculate The estimated location of the underwater AUVs in the group; the weighting coefficients are determined based on the covariance of the measurement noise of each group, and the results are applied to the neighborhood space. The estimated locations are weighted and fused to update the nonlinear measurement equations; Step 4: Perform a first-order Taylor expansion on the updated nonlinear measurement equation at the state vector prediction estimate to obtain the approximately linearized measurement equation. The information matrix increment and information vector increment are calculated, and the state vector and covariance matrix of the underwater AUV are updated based on the information filtering algorithm to output the position of the underwater AUV.

3. The underwater AUV positioning method for noise distribution modeling as described in claim 2, characterized in that: Underwater AUVs equipped with inertial measurement units The state at time is The three-axis velocity and three-axis position components are respectively and Underwater AUVs in Nonlinear state equation at time 1 for: In the formula: Here is the state transition matrix of an underwater AUV. ; This is the input matrix for an underwater AUV. ; It is a triaxial acceleration; For non-Gaussian state noise, considering that unmodeled non-Gaussian state noise reduces positioning accuracy, [the following measures are taken regarding non-Gaussian state noise]. The distribution model is as follows: In the formula: This represents the mean of the state noise; is the scaling matrix of the state noise; For the degree of freedom parameter; introduce random variables. ,exist time Represented as containing random variables The distribution is as follows: In the formula: The mean is The sum and variance are Gaussian distribution; Indicates compliance with parameters and The gamma distribution.

4. The underwater AUV positioning method for noise distribution modeling as described in claim 2, characterized in that: The reference coordinates of a surface vessel equipped with an underwater acoustic base station are as follows: While underwater AUVs equipped with underwater beacons are... The coordinates to be estimated at time 1 are Underwater beacons and underwater acoustic base stations in The slant range and horizontal angle are obtained through underwater acoustic communication at various times. and The underwater AUV equipped with a depth sensor has a depth of [missing information]. Establish a nonlinear measurement equation that can characterize the position of an underwater AUV. for: In the formula: The horizontal angle between the underwater acoustic base station and the underwater beacon; The slant distance between the underwater acoustic base station and the underwater beacon; The depth of the underwater AUV; Non-Gaussian measurement noise; considering that unmodeled non-Gaussian measurement noise reduces positioning accuracy, non-Gaussian measurement noise is... The distribution model is as follows: In the formula: To measure the mean of the noise; The scaling matrix for measuring noise; For the degree of freedom parameter; introduce random variables. ,exist time Represented as containing random variables The distribution is as follows: In the formula: The mean is The sum and variance are Gaussian distribution; Indicates compliance with parameters and The gamma distribution.

5. The underwater AUV positioning method for noise distribution modeling as described in claim 2, characterized in that: Because of the continuous movement of underwater AUVs, similar measurement values ​​will be obtained in the neighborhood space. The measured value is expressed as ,in Indicates that there is in the neighborhood space Group of measurements; based on the first set of measurements Slope distance measured in the second measurement Horizontal angle ,depth It can be solved Estimated location of the group of underwater AUVs for: The measurement equation based on neighborhood space characterizing the location of underwater AUVs can be updated as follows: In the formula: These are the weighting coefficients. For the first Covariance of the second measurement noise; updated nonlinear measurement equation and measurement noise covariance Substitute into the location solution module; for the equation In the state vector prediction estimate A first-order Taylor expansion yields an approximately linearized measurement equation: In the formula: To approximate linearize the measurement equation Regarding the estimated state vector prediction value The Jacobian matrix. This will approximate the linearization of the measurement equation. Substituting into the nonlinear measurement equation We can obtain: The information matrix increment and information vector increment used for underwater AUV positioning are as follows: In the formula: The equivalent noise covariance matrix; the state vector of the underwater AUV. and corresponding covariance matrix The updates are as follows: In the formula: , The above steps will output the estimated position of the underwater AUV. .

Images