Unmanned sailboat real-time wind field reconstruction method and system

Abstract

The invention discloses an unmanned sailboat real-time wind field reconstruction method and system, and belongs to the field of unmanned sailboat navigation and environmental awareness, the system comprises a shipborne weather station sampling module, a wind field modeling module and a data fusion and estimation module, the shipborne weather station sampling module carries out real-time collection of wind field data along with an unmanned sailboat; a space-time dynamic change model of a wind field is constructed through a wind field modeling module, finally, the state of the wind field at the next moment is predicted through a data fusion and estimation module, prediction errors are corrected in combination with wind field data collected by a shipborne weather station in real time, and measurement updating is completed. According to the method, the shipborne weather station is introduced to serve as wind field data acquisition core equipment, prediction data and real-time acquired data are fused in combination with the wind field modeling module, the method has remarkable advantages in the aspects of reconstruction precision, real-time performance and path planning efficiency, and important technical support can be provided for wide application of the unmanned sailboat in the marine environment.

Description

Technical Field

[0001] The present invention belongs to the field of unmanned sailboat navigation and environmental perception, and in particular relates to a real-time wind field reconstruction method and system for an unmanned sailboat. Background Art

[0002] The ocean wind field changes dramatically and is complex, which has a direct impact on the operation and navigation decisions of unmanned sailboats that rely on wind energy. In existing methods, most wind field information acquisition relies on meteorological models or fixed meteorological stations, while traditional methods are difficult to capture dynamic wind field characteristics in real time, especially to obtain high-precision data on wind speed and direction during the movement of unmanned sailboats.

[0003] A shipboard weather station is a mobile measuring device that integrates multiple sensors. It has been widely used in the fields of ocean monitoring and ship navigation in recent years. Compared with fixed measuring stations, shipboard weather stations can move with the ship and provide real-time and dynamic meteorological data. Typical shipboard weather stations include equipment such as ultrasonic anemometers and shutter boxes, which can not only measure wind speed and direction, but also synchronously collect temperature, humidity, air pressure and GPS location information. These functions make shipboard weather stations very suitable for real-time sampling of wind field data on unmanned sailboats, providing a new technical basis for dynamic wind field reconstruction.

[0004] In recent years, research on dynamic wind farm modeling and state estimation technology has made some progress, especially by combining the Navier-Stokes equations and Kalman filters, which can expand local measurement data to the global wind field. These methods are mainly focused on wind farm optimization, but have not been fully developed in the real-time path planning scenario of unmanned sailboats. Summary of the invention

[0005] In order to solve the shortcomings of the prior art and achieve the purpose of real-time reconstruction of the wind field of an unmanned sailboat in a dynamic environment, the present invention adopts the following technical solutions:

[0006] A method for real-time wind field reconstruction of an unmanned sailboat comprises the following steps:

[0007] Step 1: The unmanned sailboat is equipped with a ship-borne meteorological station to collect wind field data in real time. The wind field data includes wind speed and wind direction. The ship-borne meteorological station has the characteristics of mobile measurement and can collect meteorological data on the route in real time as the sailboat moves. This high temporal and spatial resolution measurement method provides rich dynamic information for wind field reconstruction. The ship-borne meteorological station also collects environmental parameters and navigation parameters.

[0008] Step 2: Wind field modeling and reconstruction;

[0009] The spatiotemporal dynamic changes of the wind field are modeled to predict the state of the wind field at the next moment. Combined with the wind field data collected in real time by the ship-borne meteorological station, the prediction error is corrected to complete the measurement update.

[0010] Furthermore, the step 2 comprises the following steps:

[0011] Step 2.1: State definition; dynamically model the wind field, divide the wind field into grid points in space, and discretize it in time;

[0012] Step 2.2: spatial discretization; spatially discretize the continuous wind field in step 2.1 to generate velocity and pressure values ​​on a discrete grid;

[0013] Step 2.3: Time discretization: The data generated in step 2.2 is time discretized to ensure the stability of the numerical solution of the model; the time derivative is discretized by the backward Euler method or other stable numerical integration methods to ensure that the numerical solution has good stability and accuracy, thereby supporting real-time updating of the wind farm model;

[0014] Step 2.4: Data fusion; predict the wind field state at the next moment based on the established wind field model Navier-Stokes, and correct the prediction error in combination with the data collected by the shipboard meteorological station. The prediction and update process is performed in an iterative manner to update the wind field estimate in real time.

[0015] Furthermore, in step 2.1, the dynamic evolution of the radial and tangential velocity components of the wind field is captured based on the Navier-Stokes equations to perform dynamic modeling of the wind field. To meet the real-time computing requirements, the main dynamic terms are retained through scale analysis, and the secondary influencing terms are ignored to simplify the Navier-Stokes equations; the main dynamic terms include the change in momentum in the vertical direction, and the complex three-dimensional equations are simplified into a computational model suitable for a two-dimensional plane, making the model calculation more efficient while retaining key wind field characteristics, thereby improving the real-time computing capability and making the equations more adaptable to the needs of dynamic wind field modeling.

[0016] Furthermore, in step 2.1, the radial and tangential wind speeds of each grid point are used as state variables, and the state variables are defined as follows:

[0017] x=[u i,j ,v i,j ] T

[0018] Among them, x is the state vector of the system, which represents the wind speed distribution of the entire grid, i and j represent the grid point index, and u i,j represents the radial wind speed component at the grid point (i, j), v i,j represents the tangential wind speed component at the grid point (i, j);

[0019] The dynamic modeling of the wind field takes into account the nonlinear changes of radial wind speed u and tangential wind speed v. The dynamic equations are as follows:

[0020]

[0021] Among them, u represents radial wind speed, v represents tangential wind speed, r, θ represent radial and angular variables of polar coordinates respectively, p represents air pressure, ρ represents air density, and v represents air motion viscosity coefficient. is the Laplace operator used to represent the diffusion term, and t represents time.

[0022] Furthermore, in step 2.2, the finite difference method is used to spatially discretize the partial derivatives in the radial and tangential directions to generate velocity and pressure values ​​at the grid points, providing a numerical calculation basis for wind field modeling;

[0023] Discretization in radial direction:

[0024]

[0025] Discretization in the tangential direction:

[0026]

[0027] Discretization of the diffusion term:

[0028]

[0029] Among them, Δr represents the spacing of the grid in the radial direction, and Δθ represents the spacing of the grid in the tangential direction. After spatial discretization, the discrete velocity and pressure values ​​at the grid points are output. The innovation of this step is that the finite difference method is used to optimize the discrete calculation, which enhances the numerical stability and spatial accuracy of the model.

[0030] Furthermore, in step 2.3, the backward Euler method is used for time discretization, and the formula is as follows:

[0031]

[0032] Among them, n represents the current time step, n+1 represents the next time step, and Δt represents the time step. After time discretization, the wind speed component of each grid point in the next time step is generated. and The innovation of this step is the use of the backward Euler method, which enhances the stability of the time integration and enables the model to maintain accuracy when experiencing high-frequency dynamic changes.

[0033] Furthermore, in step 2.4, the error between the actual wind field and the estimated wind field is recursively reduced by the unscented Kalman filter fusion algorithm, including a prediction stage and an update stage, and the formula is as follows:

[0034] Prediction stage:

[0035]

[0036] Update phase:

[0037]

[0038] in, represents the k-th wind field state vector predicted based on the k-1-th updated wind field state vector, x k-1 represents the wind field state vector updated for the k-1th time, f represents the state transfer function (given by the Navier-Stokes equation), w k represents the process noise, x k represents the k-th updated wind field state vector, K k represents the kth Kalman gain, z k represents the wind speed and direction data collected by the kth meteorological station, and h represents the measurement model. This step can output the optimized wind field estimate x k ,The innovation of improving the nonlinear processing capability of the traditional Kalman filter by UKF can significantly improve the estimation accuracy and stability.

[0039] A real-time wind field reconstruction system for an unmanned sailboat comprises a shipboard meteorological station sampling module, a wind field modeling module, and a data fusion and estimation module. According to the real-time wind field reconstruction method for an unmanned sailboat, the shipboard meteorological station sampling module collects wind field data in real time along with the unmanned sailboat, and the wind field modeling module constructs a spatiotemporal dynamic change model of the wind field. Finally, the data fusion and estimation module predicts the state of the wind field at the next moment, and combines the wind field data collected in real time by the shipboard meteorological station to correct the prediction error and complete the measurement update.

[0040] Furthermore, the wind field modeling module captures the dynamic evolution of the radial and tangential velocity components of the wind field based on the Navier-Stokes equations, performs dynamic modeling of the wind field, and in order to meet the real-time computing requirements, retains the main dynamic terms through scale analysis and ignores the secondary influencing terms to simplify the Navier-Stokes equations; the main dynamic terms include the change in momentum in the vertical direction, and simplifies the complex three-dimensional equations into a calculation model suitable for a two-dimensional plane, making the model calculation more efficient while retaining key wind field characteristics, thereby improving the real-time computing capability and making the equations more adaptable to the needs of dynamic wind field modeling.

[0041] Further, the data fusion and estimation module includes a state definition unit, a space discretization unit, a time discretization unit, and a data fusion unit;

[0042] The state definition is to divide the wind field into grid points in space and discretize it in time;

[0043] The spatial discretization is to discretize the continuous wind field in space to generate velocity and pressure values ​​on discrete grids;

[0044] The time discretization is to discretize the spatially discretized data in time to ensure the stability of the numerical solution of the model; the time derivative is discretized by the backward Euler method or other stable numerical integration methods to ensure that the numerical solution has good stability and accuracy, thereby supporting real-time updating of the wind field model;

[0045] The data fusion predicts the wind field state at the next moment based on the established wind field model Navier-Stokes, and corrects the prediction error in combination with the data collected by the ship-borne meteorological station. The prediction and update process is performed in an iterative manner to update the wind field estimate in real time.

[0046] The advantages and beneficial effects of the present invention are:

[0047] The present invention is based on wind field reconstruction technology based on real-time meteorological sampling. By integrating real-time wind field data collection of shipborne meteorological stations, Navier-Stokes equations for dynamic wind field modeling and state estimation technology, a complete technical process is formed from real-time data collection, dynamic wind field reconstruction to path optimization. The purpose is to provide efficient and accurate real-time wind field information support for unmanned sailboats for path planning and environmental monitoring, so as to ensure the efficient operation of unmanned sailboats in complex marine environments. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] Figure 1 It is a schematic diagram of the overall system architecture of an embodiment of the present invention.

[0049] Figure 2 It is a flowchart of the data fusion and estimation module executed in the system of an embodiment of the present invention.

[0050] Figure 3 It is a method flow chart of an embodiment of the present invention.

[0051] Figure 4 It is a flow chart of wind field modeling and reconstruction in the method of an embodiment of the present invention. DETAILED DESCRIPTION

[0052] The specific implementation of the present invention is described in detail below in conjunction with the accompanying drawings. It should be understood that the specific implementation described here is only used to illustrate and explain the present invention, and is not used to limit the present invention.

[0053] like Figure 1 As shown, a real-time wind field reconstruction system for an unmanned sailboat includes a ship-borne meteorological station sampling module, a wind field modeling module, and a data fusion and estimation module.

[0054] The sampling module of the shipboard weather station is the basis of this system, which is responsible for collecting meteorological data in real time during the voyage. The present invention uses the Airmar 220WX shipboard weather station, whose core components include an ultrasonic anemometer and a louvered box. The ultrasonic anemometer accurately calculates the wind speed and wind direction by measuring the difference in the propagation speed of ultrasonic waves in the air; the louvered box accurately measures the air temperature and humidity by protecting the sensor from solar radiation and other environmental interference. In addition, the weather station also integrates a pressure sensor, a GPS module and a speedometer, which can provide information such as pressure changes, position, speed and heading. These data are transmitted to the display and storage unit via cables, displayed in real time in a direct reading mode, and the instantaneous wind speed or average value is recorded at a sampling frequency of one second.

[0055] The significant advantage of the shipboard weather station sampling module is its mobile measurement capability, which can dynamically sample as the sailboat sails, not only capturing local wind field characteristics, but also forming high-temporal and spatial resolution data sets at different route locations, providing detailed raw data for wind field modeling. The high sensitivity and anti-interference of the shipboard weather station ensure the accuracy of the collected data, laying a solid foundation for the subsequent wind field modeling and reconstruction of the system.

[0056] The wind field modeling module uses the Navier-Stokes equations to accurately model the spatiotemporal dynamic changes of the wind field. The core of this module is to establish a low-order dynamic model to capture the radial and tangential changes of wind speed and direction. Specifically, the Navier-Stokes equations form a set of nonlinear partial differential equations that reflect the evolution of the wind field by describing the inertial force, viscosity, pressure gradient and external force of the fluid. This module discretizes the equations, uses polar coordinates and combines the sampling data of the shipboard meteorological station to convert complex continuous wind field problems into discrete calculation problems, thereby meeting real-time requirements.

[0057] In addition, in order to balance computational efficiency and accuracy, the model improves computational speed through reasonable simplification steps (such as ignoring minor influencing items) while retaining the main dynamic characteristics of the wind field. The module also integrates a wind field state initialization method to support the rapid generation of initial wind field states from wind speed and wind direction data collected from ship-borne meteorological stations. The wind field modeling module is the core of the entire system. It converts raw meteorological data into the spatiotemporal distribution of the wind field, providing a reliable dynamic model for subsequent data fusion and path planning.

[0058] The data fusion and estimation module uses Unscented Kalman Filter (UKF) to achieve efficient fusion of model prediction and actual measurement, and construct a dynamic estimate of the global wind field. UKF is a state estimation algorithm designed for nonlinear systems, which can recursively update the state of complex systems without relying on linearization. This module uses the Navier-Stokes model to predict the state of the wind field at the next moment, and combines the real-time measurement data of the ship-borne meteorological station to correct the prediction results, thereby reducing errors and improving estimation accuracy.

[0059] like Figure 2 As shown in the figure, the data fusion process includes two main stages: prediction stage and update stage. In the prediction stage, the system uses the model to calculate the next wind speed and wind direction state; in the update stage, the prediction deviation is corrected in combination with the meteorological station sampling data, and the wind field estimate is readjusted. In order to improve the computational efficiency, this module uses matrix operations and parallel computing technology to ensure that the UKF iteration time does not exceed 0.5 seconds. The design of this module makes the wind field reconstruction real-time and robust in a dynamic environment, providing high-precision wind field data support for sailboat navigation in complex ocean environments.

[0060] like Figure 3 As shown, a method for real-time wind field reconstruction of an unmanned sailboat comprises the following steps:

[0061] Step 1: The unmanned sailboat carries a ship-borne weather station to collect wind field data in real time;

[0062] The shipboard weather station uses the Airmar 220WX multi-function sensor as its core equipment, and its main functions are as follows:

[0063] (1) Wind speed and direction measurement: Ultrasonic anemometers are used to provide instantaneous wind speed and direction data on the route.

[0064] (2) Environmental parameter collection: The louvered box measures air temperature and humidity, and the built-in air pressure sensor obtains air pressure changes.

[0065] (3) Dynamic positioning and navigation parameter monitoring: The built-in GPS module provides the sailboat’s position, speed and heading information.

[0066] Real-time data collection and transmission: The shipboard weather station obtains data through a variety of sensors installed on the sailboat. Each sensor is connected to the central display unit through cables to realize direct reading and display of meteorological data. The collected wind speed data can record instantaneous values ​​or calculate the average value within 100 seconds, thus meeting the needs of dynamic wind field modeling.

[0067] Unlike traditional fixed weather stations, ship-borne weather stations have the feature of mobile measurement and can collect meteorological data on the route in real time as the sailboat moves. This high temporal and spatial resolution measurement method provides rich dynamic information for wind field reconstruction.

[0068] Step 2: Wind field modeling and reconstruction: Using the wind field modeling module, the Navier-Stokes equations are used to accurately model the spatiotemporal dynamic changes of the wind field, predict the state of the wind field at the next moment, and combine the wind speed and wind direction data collected in real time by the shipboard meteorological station to correct the prediction error and complete the measurement update.

[0069] The Navier-Stokes equations are a set of partial differential equations that describe fluid motion and are used to simulate the time-varying velocity and pressure fields of fluids. These equations are based on the principles of fluid dynamics and can capture the complex dynamic changes of wind speed and direction, including the effects of inertial forces, viscous forces, pressure gradients, and external forces. Although the Navier-Stokes equations are mathematically very complex, the model can be simplified through scale analysis and discretization to make it suitable for real-time calculations.

[0070] like Figure 4 As shown in FIG. 1 , in wind field reconstruction, the Navier-Stokes equations are used to capture the dynamic evolution of the radial and tangential velocity components of the wind field. Specifically, wind field modeling and reconstruction includes the following steps:

[0071] Step 2.1: State definition; Dynamic modeling of the wind field based on the Navier-Stokes equations; Divide the wind field into grid points in space and discretize them in time, with the radial and tangential velocities of each grid point as state variables.

[0072] First, the system state vector is defined as follows:

[0073] x=[u i,j ,v i,j ] T

[0074] Where i and j represent the grid point index; u i,j represents the radial wind speed component (m / s) at the grid point (i, j), v i,j represents the tangential wind speed component (m / s) of the grid point (i, j); x represents the state vector of the system, which represents the wind speed distribution of the entire grid.

[0075] The dynamic modeling of the wind farm is based on the Navier-Stokes equation, which takes into account the nonlinear changes of the radial wind speed u and the tangential wind speed v. The dynamic equation is as follows:

[0076]

[0077] Where r, θ represent the radial and angular variables of polar coordinates respectively; p represents air pressure (Pa); ρ represents air density (kg / m 3 ); v represents the air kinetic viscosity coefficient (m 2 / s); represents the Laplace operator, which is used to represent the diffusion term; t represents time.

[0078] By dividing the wind field into grid points in space and discretizing it in time, the radial and tangential velocities at each grid point are selected as state variables to form the state vector of the entire system.

[0079] Navier-Stokes equations simplified:

[0080] To meet the needs of real-time calculations, the main dynamic terms are retained through scale analysis, the secondary influencing terms are ignored, and the Navier-Stokes equations are simplified. For example, the momentum change in the vertical direction is ignored, and the complex three-dimensional equations are simplified to a calculation model suitable for a two-dimensional plane. By retaining only the main dynamic terms (ignoring the momentum change in the vertical direction) and simplifying the Navier-Stokes equations, the model calculation is made more efficient while retaining key wind field characteristics. This improves the real-time computing capability and makes the equations more adaptable to the needs of dynamic wind field modeling.

[0081] Step 2.2: Spatial discretization: spatially discretize the continuous wind field in step 2.1 to generate velocity and pressure values ​​on discrete grids; use the finite difference method to spatially discretize the partial derivatives in the radial and tangential directions to generate velocity and pressure values ​​on grid points, providing a numerical calculation basis for wind field modeling.

[0082] The specific formula is as follows:

[0083] Discretization in radial direction:

[0084]

[0085] Discretization in the tangential direction:

[0086]

[0087] Discretization of the diffusion term:

[0088]

[0089] Among them, Δr is the spacing of the grid in the radial direction, and Δθ is the spacing of the grid in the tangential direction. After spatial discretization, the discrete velocity and pressure values ​​at the grid points are output. The innovation of this step is that the finite difference method is used to optimize the discrete calculation, which enhances the numerical stability and spatial accuracy of the model.

[0090] Step 2.3: Time discretization: The data generated in step 2.2 is discretized in time to ensure the stability of the numerical solution of the model. The time derivative is discretized by the backward Euler method or other stable numerical integration methods to ensure that the numerical solution has good stability and accuracy, thereby supporting real-time updating of the wind field model.

[0091] Specifically, the backward Euler method is used for time discretization, and the formula is as follows:

[0092]

[0093] Among them, n represents the current time step, n+1 represents the next time step, and Δt represents the time step length. After time discretization, the wind speed components of each grid point in the next time step are generated. and The innovation of this step is the use of the backward Euler method, which enhances the stability of the time integration and enables the model to maintain accuracy when experiencing high-frequency dynamic changes.

[0094] Step 2.4: Data fusion; predict the wind field state at the next moment according to the Navier-Stokes model, and correct the prediction error in combination with the measurement data of the ship-borne meteorological station. The prediction and update process is performed in an iterative manner to update the wind field estimate in real time.

[0095] State estimation: Combined with the wind speed and direction data collected in real time by the shipboard weather station, the error between the actual wind field and the estimated wind field is recursively reduced through the Unscented Kalman Filter fusion algorithm (UKF). The data fusion process is completed by UKF. The algorithm flow includes the prediction stage and the update stage. The prediction stage predicts the wind field state at the next moment according to the Navier-Stokes model. The update stage corrects the prediction error and optimizes the wind field reconstruction result by combining the measured data of the shipboard weather station.

[0096] The specific formula is as follows:

[0097] Prediction stage:

[0098]

[0099] Update phase:

[0100]

[0101] Among them, x k represents the wind field state vector, represents the predicted state vector, K k is the Kalman gain, z k represents the wind speed and direction data collected by the weather station, f represents the state transfer function (given by the Navier-Stokes equation), h represents the measurement model, and wk This represents process noise.

[0102] UKF performs well in processing nonlinear systems and can significantly improve the accuracy and stability of wind field estimation.

[0103] Iterative update: At each time step, the model first predicts the state of the wind field at the next moment based on the Navier-Stokes equation; then, the prediction error is corrected in combination with the sampling data of the meteorological station to complete the measurement update. The above prediction and update process is performed iteratively to update the wind field estimate in real time.

[0104] This step can output the optimized wind field estimate x k ,The innovation of improving the nonlinear processing capability of the traditional Kalman filter by UKF can significantly improve the estimation accuracy and stability.

[0105] Through the above method, the wind field modeling module can efficiently reconstruct the wind field in a complex dynamic ocean environment. Compared with the traditional method, the reconstruction method based on the Navier-Stokes dynamic wind model has higher accuracy and dynamic adaptability, providing accurate wind speed and wind direction data support for the path planning of the unmanned sailboat. Real-time wind field reconstruction not only significantly improves the intelligent level of sailboat navigation, but also lays a technical foundation for a wider range of ocean monitoring applications.

[0106] The above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some or all of the technical features thereof may be replaced by equivalents. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A real-time wind field reconstruction method for an unmanned sailboat, characterized in that The steps include: Step 1: The unmanned sailboat carries a ship-borne weather station to collect wind field data in real time; Step 2: Wind field modeling and reconstruction; The spatiotemporal dynamic changes of the wind field are modeled to predict the state of the wind field at the next moment. Combined with the wind field data collected in real time by the ship-borne meteorological station, the prediction error is corrected to complete the measurement update.

2. The method for real-time wind field reconstruction of an unmanned sailboat according to claim 1, characterized in that: The step 2 comprises the following steps: Step 2.1: State definition; dynamically model the wind field, divide the wind field into grid points in space, and discretize it in time; Step 2.2: spatial discretization; spatially discretize the continuous wind field in step 2.1 to generate velocity and pressure values ​​on a discrete grid; Step 2.3: Time discretization: Time discretize the data generated in step 2.2; Step 2.4: Data fusion; predict the wind field state at the next moment based on the established wind field model, and correct the prediction error in combination with the data collected by the ship-borne meteorological station. The prediction and update process is performed in an iterative manner to update the wind field estimate in real time.

3. The method for real-time wind field reconstruction of an unmanned sailboat according to claim 2, characterized in that: In step 2.1, the dynamic evolution of the radial and tangential velocity components of the wind field is captured based on the Navier-Stokes equations, and dynamic modeling of the wind field is performed. The main dynamic terms are retained through scale analysis, and the secondary influencing terms are ignored to simplify the Navier-Stokes equations; the main dynamic terms include the change in momentum in the vertical direction, and the three-dimensional equations are simplified into a calculation model suitable for a two-dimensional plane.

4. The method for real-time wind field reconstruction of an unmanned sailboat according to claim 2, characterized in that: In step 2.1, the radial and tangential wind speeds of each grid point are taken as state variables, and the state variables are defined as follows: x=[u i,j ,v i,j ] T Among them, x is the state vector of the system, which represents the wind speed distribution of the entire grid, i and j represent the grid point index, and u i,j represents the radial wind speed component at the grid point (i, j), v i,j represents the tangential wind speed component at the grid point (i, j); The dynamic modeling equation of the wind field is as follows: Among them, u represents radial wind speed, v represents tangential wind speed, r, θ represent radial and angular variables of polar coordinates respectively, p represents air pressure, ρ represents air density, and v represents air motion viscosity coefficient. is the Laplace operator used to represent the diffusion term, and t represents time.

5. The method for real-time wind field reconstruction of an unmanned sailboat according to claim 4, characterized in that: In step 2.2, the partial derivatives in the radial and tangential directions are spatially discretized using the finite difference method to generate velocity and pressure values ​​at the grid points; Discretization in radial direction: Discretization in the tangential direction: Discretization of the diffusion term: Where Δr represents the spacing of the grid in the radial direction, and Δθ represents the spacing of the grid in the tangential direction.

6. The method for real-time wind field reconstruction of an unmanned sailboat according to claim 4, characterized in that: In step 2.3, the backward Euler method is used for time discretization, and the formula is as follows: Among them, n represents the current time step, n+1 represents the next time step, and Δt represents the time step. After time discretization, the wind speed component of each grid point in the next time step is generated. and 7. The method for real-time wind field reconstruction of an unmanned sailboat according to claim 2, characterized in that: In step 2.4, the error between the actual wind field and the estimated wind field is recursively reduced by the unscented Kalman filter fusion algorithm, including the prediction stage and the update stage, and the formula is as follows: Prediction stage: Update phase: in, represents the k-th wind field state vector predicted based on the k-1-th updated wind field state vector, x k-1 represents the wind field state vector updated for the k-1th time, f represents the state transfer function, and w k represents the process noise, x k represents the k-th updated wind field state vector, K k represents the kth Kalman gain, z k represents the data collected by the kth weather station, and h represents the measurement model.

8. A real-time wind field reconstruction system for an unmanned sailboat, comprising a shipboard meteorological station sampling module, a wind field modeling module, and a data fusion and estimation module, characterized in that: According to the real-time wind field reconstruction method of an unmanned sailboat as described in claim 1, the wind field data is collected in real time along with the unmanned sailboat through the ship-borne meteorological station sampling module, and the spatiotemporal dynamic change model of the wind field is constructed through the wind field modeling module. Finally, the state of the wind field at the next moment is predicted through the data fusion and estimation module, and the wind field data collected in real time by the ship-borne meteorological station is combined to correct the prediction error and complete the measurement update.

9. The unmanned sailboat real-time wind field reconstruction system according to claim 8, characterized in that: The wind field modeling module captures the dynamic evolution of the radial and tangential velocity components of the wind field based on the Navier-Stokes equations, performs dynamic modeling of the wind field, retains the main dynamic terms through scale analysis, and ignores the secondary influencing terms to simplify the Navier-Stokes equations; the main dynamic terms include the change in momentum in the vertical direction, and the three-dimensional equations are simplified into a calculation model suitable for a two-dimensional plane.

10. The unmanned sailboat real-time wind field reconstruction system according to claim 8, characterized in that: The data fusion and estimation module includes a state definition unit, a space discretization unit, a time discretization unit, and a data fusion unit; The state definition is to divide the wind field into grid points in space and discretize it in time; The spatial discretization is to discretize the continuous wind field in space to generate velocity and pressure values ​​on discrete grids; The time discretization is to perform time discretization on the spatially discretized data; The data fusion predicts the wind field state at the next moment based on the established wind field model, and corrects the prediction error in combination with the data collected by the ship-borne meteorological station. The prediction and update process is performed in an iterative manner to update the wind field estimate in real time.

Images