CN-121994276-A - Deep sea route deviation early warning method and system based on USBL

CN121994276ACN 121994276 ACN121994276 ACN 121994276ACN-121994276-A

Abstract

The invention discloses a deep sea route deviation early warning method and system based on USBL, wherein the method comprises the steps of carrying out installation deviation calibration and sound velocity profile calibration on a USBL transducer array, and executing time synchronization on all sensors; the method comprises the steps of collecting original data of an underwater target, carrying out outlier detection and signal filtering, calculating the relative azimuth of the target based on the phase difference of an USBL transducer array, calculating the distance of the target through sound wave round trip time, converting the position in a USBL coordinate system into absolute longitude and latitude coordinates by combining IMU gesture data and GPS position data, calculating vertical deviation, prospective deviation and depth deviation through a route deviation detection algorithm, optimizing deviation calculation accuracy by combining an adaptive filtering algorithm, an installation deviation joint calibration algorithm and a multi-source data fusion result, generating corresponding early warning information and starting a system adaptive adjustment mechanism. According to the method, the early warning and accurate monitoring of the routing deviation of the underwater vehicle are realized by reducing redundant operation in the deviation calculation process and improving the data processing precision.

Inventors

DU PENG
XIN SHENG
WANG CHUNXING
WU QIBAO
QIAO YUEKUN

Assignees

深海智人(广州)技术有限公司

Dates

Publication Date: 20260508
Application Date: 20251231

Claims (10)

1. A USBL-based deep sea route deviation early warning method is characterized by comprising the following steps: Performing installation deviation calibration on the USBL transducer array, performing sound velocity profile calibration on the USBL transducer array through the CTD sensor, and performing time synchronization on all the sensors; Collecting original data of an underwater target, wherein the original data comprises USBL acoustic signals, IMU attitude data, GPS position data and depth data, and performing outlier detection and signal filtering on the original data; Calculating the relative azimuth of a target based on the phase difference of the USBL transducer array, calculating the distance of the target through the sound wave round trip time, and converting the position in the USBL coordinate system into absolute longitude and latitude coordinates by combining IMU gesture data and GPS position data; calculating vertical deviation, prospective deviation and depth deviation through a route deviation detection algorithm, and optimizing deviation calculation accuracy by combining an adaptive filtering algorithm, an installation deviation joint calibration algorithm and a multi-source data fusion result; and evaluating the deviation grade, generating corresponding early warning information, displaying the early warning information through a human-computer interaction interface, and starting a system self-adaptive adjustment mechanism.
2. The method of claim 1, wherein said calibrating the USBL transducer array for installation bias, calibrating the sound velocity profile through CTD sensors, performing time synchronization for all sensors, comprises: The USBL transducer array is fixed at the bottom of a mother ship through a special mounting bracket and is rigidly connected with the ship, the three-dimensional lever arm deviation between the transducer array and a GPS antenna of the mother ship is measured, the splayed or circular track calibration maneuver is executed in a calm sea area, multi-source data are collected, and the mounting deviation parameter is calculated through a joint calibration algorithm; Collecting temperature, salinity and pressure data of different depths through a CTD sensor, calculating a real-time sound velocity profile by utilizing a Wilson sound velocity formula, and establishing a sound ray bending correction model; microsecond synchronization of all sensors is achieved through a PTP precision clock protocol, and data acquisition frequencies are set uniformly.
3. The method of claim 1, wherein the calculating vertical deviation, look-ahead deviation, and depth deviation by the route deviation detection algorithm comprises: First discretizing a predetermined route into a series of successive route segments, each route segment being composed of two successive route points And Definition; For any instant measured position of underwater vehicle The algorithm calculates its route segment by the following mathematical method Is defined by the vertical deviation of: routing segment Any point above can be expressed as: ; Point(s) To route segment Corresponding parameters to the closest point of (a): ; If it is The closest point is If (3) The closest point is Otherwise, the nearest point is ; The perpendicular distance from point Q to the route segment is: ; in addition to the vertical deviation, the algorithm also calculates a look-ahead deviation, predicting the trend of the positional deviation over a period of time in the future: ; Wherein the method comprises the steps of Is a velocity vector of the underwater equipment; taking measurement noise and system errors into consideration, an error model based on the Liqun theory is adopted by the algorithm, and the navigation error is defined as a right group error: ; Wherein the method comprises the steps of For the purpose of an attitude error, In order to be a speed error, Is a position error.
4. The method of claim 1, wherein the adaptive filtering algorithm is an Adaptive Robust Unscented Kalman Filtering (ARUKF) algorithm based on vertical constraints, comprising: State vector definitions, state vectors contain position, speed, attitude and sensor error parameters of the underwater vehicle: ; Wherein the method comprises the steps of As a function of the position of the object, In order to be able to achieve a speed, In order for the euler angle to be the value, For the purpose of biasing the accelerometer, Biasing the gyroscope; unscented transformation, using the nonlinear characteristics of Unscented Kalman Filter (UKF) processing systems, propagating state distributions through a set of carefully selected Sigma points; Robust estimation, by introducing an equivalent weight function, the weight of an abnormal observed value is reduced: ; Wherein the method comprises the steps of In order to normalize the residual error, the residual error is normalized, For the empirical threshold, take ; Vertical constraints, in combination with accurate depth information provided by the depth sensor, impose vertical constraints: ; Where Z USBL is the USBL measurement depth and Z depth-sensor is the depth sensor measurement depth.
5. The method of claim 1, wherein the installation deviation joint calibration algorithm comprises: The observation equation is established, and the observation equation taking the installation deviation into consideration can be expressed as: ; Wherein the method comprises the steps of For the USBL observations of the values of the USBL, In order to rotate the matrix is rotated, As a result of the true position, For the lever arm deflection, Is observation noise; And establishing a joint adjustment equation by adopting a Gaussian-Markov model: ; and (3) parameter estimation, namely solving the joint model parameters based on a least square principle: ; Wherein the method comprises the steps of For an optimal estimate of a parameter to be estimated, the parameter to be estimated comprising an installation deviation and an acoustic velocity error, As a global parameter of the system, Is a local parameter.
6. The method of claim 1, wherein the multi-source data fusion is implemented by federal kalman filtering, and wherein inertial navigation, doppler and USBL positioning information are fused to output optimized state vector data for position, velocity, attitude and sensor errors, providing input for route bias calculation.
7. The method of claim 1, wherein the bias level comprises: Normal level, wherein the horizontal deviation is less than 2.5m, the vertical deviation is less than 1.0m, and the system monitors and records data normally; A focus level, namely, horizontal deviation is 2.5m-5.0m, vertical deviation is 1.0m-2.0m, and the system marks the deviation and focuses on the trend; warning level, wherein the horizontal deviation is 5.0m-10.0m, the vertical deviation is 2.0m-4.0m, and the system gives out acousto-optic early warning; Critical level-horizontal deviation >10.0m and vertical deviation >4.0m, the system issues emergency alerts and recommends immediate intervention.
8. The deep sea route deviation early warning system based on the USBL is characterized by comprising a hardware perception layer, a data processing layer and an early warning application layer which are sequentially connected in a communication manner and used for executing the method of any one of claims 1-7; The hardware perception layer is used for acquiring original physical data of an underwater target and comprises a USBL transducer array, a high-precision Inertial Measurement Unit (IMU), a depth sensor, a sound velocity profiler and a GPS receiving module; the data processing layer is used for processing the original data and calculating real-time route deviation and comprises a signal processing module, a route deviation calculating module, a self-adaptive filtering module, an installation deviation joint calibration module and a multi-source data fusion module; the early warning application layer is used for generating hierarchical early warning and supporting decision intervention and comprises a deviation evaluation module, an early warning generation module, a man-machine interaction interface and a self-adaptive adjustment module.
9. The system of claim 8, wherein the USBL transducer array is designed with a circular array layout and multi-frequency coded signals, the circular array is formed by arranging piezoelectric ceramic sensor units at equal intervals, the multi-frequency coded signals are designed with adjustable frequency and pseudo-random coded modulation, and each sensor unit is connected with a low-noise preamplifier and a high-precision ADC analog-to-digital converter; the high-precision Inertial Measurement Unit (IMU) is rigidly connected with the USBL transducer array by adopting a combination of a fiber optic gyroscope and a quartz accelerometer; the sound velocity profile meter is a CTD sensor; The depth sensor is a silicon piezoresistive pressure sensor; the GPS receiving module is a double-frequency GPS receiver.
10. The system of claim 8, wherein the route bias calculation module is configured with a route bias detection algorithm, the adaptive filter module is configured with a ARUKF algorithm, and the installation bias joint calibration module is configured with an installation bias joint calibration algorithm; the man-machine interaction interface comprises a routing plan, a deviation trend graph, an early warning state panel and a system state monitoring module; the self-adaptive adjustment module comprises an environment self-adaptive mechanism, a sensor fault detection and isolation mechanism and an early warning sensitivity self-adaptive mechanism, and is used for dynamically updating the sound velocity profile, isolating a fault sensor and adjusting an early warning threshold value.

Description

Deep sea route deviation early warning method and system based on USBL Technical Field The invention relates to the technical field of deep sea navigation and underwater positioning, in particular to a deep sea route deviation early warning method and system based on USBL. Background In the fields of deep sea development, underwater exploration and the like, accurate routing control of underwater vehicles (such as AUV and ROV) is a core link for ensuring task success. The underwater equipment needs to execute tasks such as detection, operation and the like according to a preset route, and the position positioning and track calibration of the underwater equipment mainly depend on an underwater sound positioning system, wherein an ultra-short baseline (USBL) system becomes one of main technologies of underwater positioning due to flexible deployment and moderate cost. In the prior art, the typical implementation flow of the underwater navigation positioning system is that the underwater equipment carries a sensor to acquire the position information of the underwater equipment, the underwater equipment interacts with a mother ship or a beacon to finish positioning, and a navigation algorithm is combined to adjust the navigation state. For example, in the prior art, a split AUV underwater navigation positioning system realizes the position acquisition of an underwater robot by integrating a navigation sensor and a positioning module, and provides data support for subsequent navigation control. However, the deep sea environment has specificity and complexity, and in practical application, various technical bottlenecks exist, namely, on one hand, the factors such as sound ray bending, water flow interference and the like in the marine environment and unavoidable installation deviation in the equipment installation process can cause deviation between the actual track of the underwater equipment and a preset route, and on the other hand, the existing USBL system and related navigation positioning scheme can only realize acquisition and output of position information and lack a real-time detection, trend prejudgment and grading early warning mechanism special for the route deviation. The defect directly causes that the prior art cannot timely send out early warning at the initial stage of deviation occurrence, and cannot provide decision intervention basis for operators, so that risks such as task interruption, equipment collision damage and the like are caused. The method is mainly characterized in that the prior art focuses on the basic function of 'position acquisition', deviation analysis and early warning logic are not designed aiming at the dynamic interference characteristics of the deep sea environment, the requirements of positioning accuracy and intervention timeliness are difficult to balance, and the reliability of routing control of the underwater equipment is insufficient. Disclosure of Invention The embodiment of the invention provides a deep sea route deviation early warning method and a system based on USBL, which realize early warning and accurate monitoring of the underwater vehicle route deviation, improve the reliability and safety of deep sea navigation positioning and solve the problem that the prior art lacks a special route deviation detection, pre-judging and grading early warning mechanism. In order to achieve the above purpose, the present invention provides the following technical solutions: a deep sea route deviation early warning method based on USBL includes: Performing installation deviation calibration on the USBL transducer array, performing sound velocity profile calibration on the USBL transducer array through the CTD sensor, and performing time synchronization on all the sensors; Collecting original data of an underwater target, wherein the original data comprises USBL acoustic signals, IMU attitude data, GPS position data and depth data, and performing outlier detection and signal filtering on the original data; Calculating the relative azimuth of a target based on the phase difference of the USBL transducer array, calculating the distance of the target through the sound wave round trip time, and converting the position in the USBL coordinate system into absolute longitude and latitude coordinates by combining IMU gesture data and GPS position data; calculating vertical deviation, prospective deviation and depth deviation through a route deviation detection algorithm, and optimizing deviation calculation accuracy by combining an adaptive filtering algorithm, an installation deviation joint calibration algorithm and a multi-source data fusion result; and evaluating the deviation grade, generating corresponding early warning information, displaying the early warning information through a human-computer interaction interface, and starting a system self-adaptive adjustment mechanism. Further, the performing the installation deviation calibration on the USBL transducer array, perfor