Adaptive path tracking control method and device for unmanned ship

Abstract

The present invention belongs to the technical field of unmanned ship tracking and control, and discloses an unmanned ship adaptive path tracking control method and device; the method includes: real-time monitoring of position data through a boundary judgment algorithm to ensure that the ship is always in a preset safe area; synchronously collecting multi-source data such as ocean currents, radiation, and batteries, constructing three major constraints for predicting drift trajectory, charging area, and remaining endurance time, and using a path planning algorithm to dynamically generate a global optimal path to balance the shortest distance, low energy consumption, and high charging benefits. For energy management, preset multi-level power thresholds trigger differentiated loss reduction strategies. At the control layer, the propeller control parameters are calculated in real time based on path deviation, heading deviation, and environmental interference; the present invention improves the autonomous navigation capability, energy utilization efficiency, and mission sustainability of unmanned ships in dynamic environments through the coordinated optimization of global path planning and local control, and is suitable for scenarios such as ocean monitoring and resource exploration.

Description

Technical Field

[0001] The present invention relates to the technical field of unmanned vessel tracking control, and more particularly to an unmanned vessel adaptive path tracking control method and device. Background Art

[0002] The Pacific Ocean is Earth's largest ocean, accounting for approximately 49.8% of the Earth's total ocean surface. It has a profound impact on global climate, ecosystems, economic and social development. It is not only a vital pathway for the transport of heat and water vapor globally, but also plays a key role in regulating global climate and maintaining biodiversity. However, due to the vastness and complex environment of the Pacific Ocean, traditional ocean observation methods such as manned vessels, satellite remote sensing, and buoy monitoring have many limitations.

[0003] Traditional observation vessels, limited by manpower, endurance, and adverse weather conditions, struggle to conduct sustained observations over extended periods of time and over large areas. Satellite remote sensing, while capable of providing observation data over large areas, is limited in its ability to obtain information below sea level and is susceptible to interference from factors such as cloud cover and sea conditions. Buoy monitoring, on the other hand, is fixed in specific locations, resulting in a narrow data coverage area and a failure to fully reflect the dynamic changes in the marine environment.

[0004] World Patent Application Publication No. WO2021035911A1 discloses a method and system for unmanned vehicle path planning based on a data-driven linear variable parameter genetic algorithm. The system receives navigation data, ship logs, and status information from the unmanned vehicle; converts the navigation data into a coordinate system and plots a coordinate graph, marking the target point and the real-time position of the unmanned vehicle on the graph; and uses a data-driven linear variable parameter genetic algorithm to generate the optimal trajectory for the unmanned vehicle path planning. This invention overcomes the shortcomings of traditional genetic algorithms, such as slow convergence, low computational efficiency, and susceptibility to premature convergence (local optimum).

[0005] Although the above method can meet most scenarios, research and practical application of the above method and existing technology have found that the above method and existing technology have at least the following defects:

[0006] In a complex and dynamic ocean environment, it is impossible to achieve path planning and precise control of unmanned ships under multiple dynamic constraints.

[0007] In view of this, the present invention proposes an unmanned vessel adaptive path tracking control method and device to solve the above problems. Summary of the Invention

[0008] In order to overcome the above-mentioned defects of the prior art and to achieve the above-mentioned purpose, the present invention provides the following technical solution: an unmanned ship adaptive path tracking control method, comprising the following steps:

[0009] Collect the position data of the unmanned ship, and use the boundary judgment algorithm to determine whether the unmanned ship exceeds the preset boundary based on the position data of the unmanned ship. If it exceeds, proceed to the next step, otherwise continue to judge;

[0010] Collect and analyze ocean current data, environmental data, and unmanned vessel data corresponding to the location data to obtain the predicted drift trajectory of the unmanned vessel; collect and analyze radiation data corresponding to the location data to obtain the charging area; collect and analyze battery data to obtain the predicted remaining battery life; construct a path planning target based on the predicted drift trajectory, charging area, and predicted remaining battery life, and use the path planning algorithm to perform dynamic path search to obtain the global optimal path;

[0011] Preset multiple power thresholds. When the current remaining battery power in the battery data falls below different thresholds, the corresponding loss reduction strategy is activated.

[0012] The path deviation and heading deviation are calculated based on the position data and the global optimal path. The path deviation, heading deviation, ocean current data, environmental data, battery data and the global optimal path are used as inputs of the path tracking model to obtain the unmanned ship propulsion control parameters; real-time adjustments are made based on the unmanned ship propulsion control parameters to achieve path tracking control.

[0013] Furthermore, the method for obtaining the predicted drift trajectory of the unmanned vessel includes:

[0014] According to the ocean current data, environmental data and unmanned ship data corresponding to the position data, based on Newton's laws of motion and fluid dynamics, combined with the ocean current data, environmental data and unmanned ship data corresponding to the position data, the motion equation of the unmanned ship under external force is established, the acceleration of the unmanned ship is extracted from the unmanned ship data, and the position and velocity of the unmanned ship in the next time step are calculated by combining the velocity-acceleration formula and the displacement-velocity formula; the position sequence within the preset trajectory period is obtained through iterative calculation, and the position sequence is used as the predicted drift trajectory of the unmanned ship.

[0015] Furthermore, the method for obtaining the charging area includes:

[0016] Obtain the UTC time corresponding to the collected unmanned vessel's position data, convert the UTC time to Julian day, calculate the Greenwich sidereal time based on the Julian day, and calculate the local sidereal time based on the Greenwich sidereal time;

[0017] Calculate the mean ecliptic longitude and the correction term for the Earth's orbit eccentricity based on the Julian day, calculate the solar ecliptic longitude based on the mean ecliptic longitude and the correction term for the Earth's orbit eccentricity, calculate the solar right ascension based on the solar ecliptic longitude, and obtain the hour angle based on the local sidereal time and the solar right ascension;

[0018] Calculate the solar declination angle through the obliquity of the ecliptic and the solar ecliptic longitude;

[0019] Based on the longitude and latitude coordinates of the unmanned vessel, combined with the hour angle and solar declination angle, the corresponding solar altitude angle is calculated, and the solar altitude angle sequence in the preset future period is obtained through iterative calculation;

[0020] Select the period in the solar altitude angle sequence when the solar altitude angle is greater than the preset altitude angle threshold as the optimal charging period of the day;

[0021] Collecting illumination data from the solar panel locations on the unmanned vessel. The illumination data includes the illumination sequence and cloud movement speed within a preset time period. This illumination data is used as input to a short-term prediction model to obtain the predicted illumination intensity within a preset future time period.

[0022] Obtain satellite images from radiation data, identify cloud pixels in the image through threshold segmentation, calculate the proportion of cloud area in the preset sea area, and mark it as cloud cover rate;

[0023] Obtain the cloud movement speed from the illumination data, use the cloud coverage and cloud movement speed as inputs to the cloud prediction model, and obtain the cloud coverage sequence within a preset future period;

[0024] If the predicted light intensity is greater than the preset light intensity threshold and the cloud coverage in the preset future time period is lower than the preset coverage threshold, the corresponding solar panel location is marked as a charging area, and the optimal charging time of the day is marked on the charging area.

[0025] Furthermore, the method for obtaining the predicted remaining battery life includes:

[0026] Obtain the current remaining power from the battery data, as well as the propeller power and mission equipment power consumption. Correct the propeller power according to the ocean current speed and the speed of the unmanned vessel to obtain the propeller power correction value. Calculate the total energy consumption based on the propeller power correction value and the mission equipment power consumption.

[0027] Obtain the output power, incident light power and panel area of the solar panel to obtain the conversion efficiency of the solar panel;

[0028] The charge capacity corresponding to the current moment is calculated based on the light intensity and the conversion efficiency of the solar panel. The charge capacity at different moments is calculated based on the predicted light intensity and the conversion efficiency of the solar panel in the preset future period. The charge capacity at different moments in the preset future period is accumulated to obtain the charge capacity in the preset future period.

[0029] The remaining battery life is calculated based on the current remaining battery power, the charge level in the preset future period, and the total energy consumption.

[0030] Furthermore, the method for obtaining the global optimal path includes:

[0031] Obtain charging areas, drift risk areas, range restricted areas, and preset optimization targets;

[0032] The drift risk zone is where the predicted ocean current speed exceeds the preset ocean current speed threshold, and the endurance restricted zone is where the total route time exceeds the remaining endurance time. The preset optimization goals include minimizing total energy consumption, maximizing charging benefits, and ensuring that the total route time is less than the remaining endurance time.

[0033] Calculate the Euclidean distance between the location data and the charging area; calculate the ratio of the electrical energy ultimately stored in the battery to the total energy of the incident light to obtain the energy conversion efficiency; calculate the product of the light intensity, the effective area of the solar panel, and the nominal efficiency of the solar panel to obtain the input power of the solar panel; calculate the product of the energy conversion efficiency, the input power of the solar panel, and the effective charging time to obtain the charging benefit; and weight the Euclidean distance, energy conversion efficiency, and charging benefit to calculate the cost function.

[0034] A heuristic function is constructed based on the Euclidean distance and the weighted attractiveness of the charging area, where the weight of the attractiveness of the charging area is a negative value. By minimizing the cost function and combining the heuristic function to expand the nodes, the global optimal path is obtained.

[0035] Furthermore, the charging area attraction is obtained by summing the attractions of all charging areas, and the attraction is the ratio of the light intensity corresponding to the charging area to the Euclidean distance.

[0036] Furthermore, in the method for obtaining the global optimal path, trigger conditions for dynamic path adjustment are also preset. When the trigger conditions are met, the global optimal path is replanned; the trigger conditions include: the deviation value between the actual drifting trajectory and the predicted drifting trajectory is greater than a preset deviation threshold; the predicted remaining endurance time is lower than a preset endurance safety threshold.

[0037] Furthermore, the method for determining whether the unmanned ship exceeds a preset boundary based on the position data of the unmanned ship based on the boundary judgment algorithm includes:

[0038] Preset the boundaries of the observation area and pre-calibrate the boundary coordinates of the observation area through GIS;

[0039] The current longitude and latitude coordinates of the unmanned ship are obtained through GIS, and the Euclidean distance between the current longitude and latitude coordinates of the unmanned ship and the boundary coordinates is calculated in real time. If the Euclidean distance exceeds the preset distance threshold, the path replanning is triggered immediately.

[0040] Furthermore, the method of starting the corresponding loss reduction strategy includes:

[0041] The first, second, and third thresholds are preset. When the battery level drops below the first threshold, the preset secondary tasks are suspended. When the battery level drops below the second threshold, only the preset core tasks are retained. When the battery level drops below the third threshold, the aircraft is forced to return home or enter sleep mode, with the return route planned to prioritize the charging area.

[0042] Path deviation, heading deviation, ocean current data, environmental data, battery data and the global optimal path are used as inputs of the path tracking model to obtain the unmanned ship propulsion control parameters; real-time adjustments are made based on the unmanned ship propulsion control parameters to achieve path tracking control.

[0043] Furthermore, the method for implementing path tracking control includes:

[0044] Based on the position data of the unmanned vessel, the difference between the actual position and the nearest point of the preset path is calculated to obtain the path deviation; the difference between the current heading and the preset heading is calculated to obtain the heading deviation;

[0045] Path deviation, heading deviation, ocean current data, environmental data, battery data and the global optimal path are used as inputs of the path tracking model to obtain the unmanned ship propulsion control parameters; real-time adjustments are made based on the unmanned ship propulsion control parameters to achieve path tracking control.

[0046] An unmanned ship adaptive path tracking control device implements the unmanned ship adaptive path tracking control method, comprising:

[0047] Boundary judgment module: collects the position data of the unmanned ship and uses the boundary judgment algorithm to determine whether the unmanned ship exceeds the preset boundary based on the position data of the unmanned ship. If it exceeds, it proceeds to the next step, otherwise it continues to judge;

[0048] Path planning module: collects and analyzes ocean current data, environmental data, and unmanned vessel data corresponding to the location data to obtain the predicted drift trajectory of the unmanned vessel; collects and analyzes radiation data corresponding to the location data to obtain the charging area; collects and analyzes battery data to obtain the predicted remaining battery life; constructs a path planning target based on the predicted drift trajectory, charging area, and predicted remaining battery life, and uses the path planning algorithm to perform dynamic path search to obtain the global optimal path;

[0049] Adaptive loss reduction module: preset multiple power thresholds. When the current remaining battery power in the battery data falls below different thresholds, the corresponding loss reduction strategy is activated;

[0050] Path tracking module: Calculates path deviation and heading deviation based on position data and the global optimal path, uses path deviation, heading deviation, ocean current data, environmental data, battery data, and the global optimal path as inputs to the path tracking model to obtain the propulsion control parameters of the unmanned vessel; performs real-time adjustments based on the propulsion control parameters of the unmanned vessel to achieve path tracking control.

[0051] The technical effects and advantages of the unmanned ship adaptive path tracking control method and device of the present invention are as follows:

[0052] The present invention monitors the position of the unmanned ship in real time through a boundary judgment algorithm to ensure that it is always within a preset safe area; at the same time, it integrates ocean current data, radiation data and environmental data, dynamically predicts the drift trajectory and identifies the charging area, and combines the remaining battery power with the multi-level power threshold strategy, etc., and adopts the A* algorithm combined with the Dijkstra algorithm to optimize the path planning, balance the shortest path and charging benefits to extend the endurance; it also optimizes the propeller control parameters using the path tracking model through real-time feedback of path deviation and heading deviation, reduces navigation errors, and automatically triggers the loss reduction strategy when the battery is low; the present invention solves the problem of path planning and precise control of unmanned ships in an open ocean environment under multiple dynamic constraints through the coordinated optimization of global path planning and local control, significantly improving the autonomous navigation capability, energy utilization efficiency and mission sustainability of unmanned ships in dynamic environments, and is suitable for scenarios such as ocean monitoring and resource exploration. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] Figure 1 Schematic diagram of the adaptive path tracking control method for an unmanned ship according to embodiment 1 of the present invention;

[0054] Figure 2 This is a schematic diagram of the solar panel installation according to Example 1 of the present invention;

[0055] Figure 3 This is a schematic diagram of the method flow of Example 2 of the present invention;

[0056] Figure 4 This is a schematic structural diagram of an adaptive path tracking control device for an unmanned ship according to embodiment 3 of the present invention. DETAILED DESCRIPTION

[0057] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0058] Example 1

[0059] See also Figure 1 As shown, the unmanned ship adaptive path tracking control method described in this embodiment includes the following steps:

[0060] The UAV collects its position data and uses a boundary judgment algorithm to determine whether it has exceeded the preset boundary. If so, path planning begins; otherwise, continued judgment continues. Position data includes longitude and latitude coordinates, heading angle, pitch angle, and roll angle. Collecting the UAV's position data and using the boundary judgment algorithm to monitor whether it has exceeded the preset boundary in real time is the core safety guarantee for the UAV's adaptive path tracking control. The boundary judgment algorithm continuously compares the actual position with the geofence coordinates. Once a boundary violation is detected, dynamic path replanning (such as launching the A* algorithm to re-search for a safe path) is triggered. The algorithm also adjusts the heading of the propeller control parameters to forcibly pull the UAV back to a safe area. Boundary judgment also provides constraints for path planning, ensuring that the globally optimal path always stays within legal waters and preventing the algorithm from exceeding safety thresholds due to blind pursuit of efficiency. Furthermore, this function works in conjunction with the energy management module. When a charging area is available in the direction of the boundary violation, the path is prioritized to balance safety and endurance requirements.

[0061] The method for determining whether the unmanned ship exceeds the preset boundary based on the position data of the unmanned ship based on the boundary judgment algorithm includes:

[0062] Preset the boundaries of the observation area and pre-calibrate the boundary coordinates of the observation area through GIS;

[0063] The current longitude and latitude coordinates of the unmanned ship are obtained through GIS, and the Euclidean distance between the current longitude and latitude coordinates of the unmanned ship and the boundary coordinates is calculated in real time. If the Euclidean distance exceeds the preset distance threshold, the path replanning is triggered immediately.

[0064] Collect and analyze ocean current data, environmental data, and unmanned vessel data corresponding to the location data to obtain the predicted drift trajectory of the unmanned vessel; collect and analyze radiation data corresponding to the location data to obtain the charging area; collect and analyze battery data to obtain the predicted remaining battery life; construct a path planning target based on the predicted drift trajectory, charging area, and predicted remaining battery life, and use the path planning algorithm to perform dynamic path search to obtain the global optimal path;

[0065] Methods for obtaining the predicted drift trajectory of the unmanned vessel include:

[0066] Based on the ocean current data corresponding to the position data, such as ocean current velocity and direction, environmental data such as wind speed, wind direction, wave height, wave direction and wave period, and unmanned ship data such as acceleration, angular velocity and attitude angle of the ship, the motion equation of the unmanned ship under the action of external forces is established based on Newton's laws of motion and fluid dynamics, combined with the ocean current data, environmental data and unmanned ship data corresponding to the position data, such as ; ; ;in, For the quality of unmanned ships; is the hull velocity vector; For time; is the thrust of the propeller, which is calculated based on the product of the acceleration of the hull and the mass of the unmanned ship; For ocean current force; For wind power; is the wave force; For resistance; is the density of seawater; is the displacement volume of the hull; is the ocean current speed; is the drag coefficient; is the frontal area of the hull; is the actual acceleration of the unmanned ship; the acceleration of the unmanned ship is extracted from the unmanned ship data, and the position and velocity of the unmanned ship in the next time step are calculated by combining the velocity-acceleration formula and the displacement-velocity formula; the position sequence within the preset trajectory period is obtained through iterative calculation, and the position sequence is used as the predicted drift trajectory of the unmanned ship.

[0067] By real-time monitoring of ocean current speed, direction, and meteorological conditions (such as wind speed and wave height), combined with the unmanned vessel's current state (such as heading and propeller power), the system can construct a high-precision drift model and predict positional deviations in future time periods. This prediction provides key input for path planning: on the one hand, it allows for early avoidance of high-risk areas (such as rapids) in global path search, optimizing path energy consumption; on the other hand, it compensates for ocean current interference in real time during local control, adjusting propeller control parameters and reducing the deviation between the actual navigation trajectory and the planned path. Furthermore, the predicted drift trajectory can be coordinated with a boundary judgment algorithm. When a boundary crossing is detected in the future, path replanning is triggered in advance to ensure navigation safety.

[0068] Methods for obtaining charging areas include:

[0069] Obtain the UTC time corresponding to the collected unmanned vessel's position data, convert the UTC time to Julian day, calculate the Greenwich sidereal time based on the Julian day, and calculate the local sidereal time based on the Greenwich sidereal time;

[0070] Calculate the mean ecliptic longitude and the correction term for the Earth's orbit eccentricity based on the Julian day, calculate the solar ecliptic longitude based on the mean ecliptic longitude and the correction term for the Earth's orbit eccentricity, calculate the solar right ascension based on the solar ecliptic longitude, and obtain the hour angle based on the local sidereal time and the solar right ascension;

[0071] Calculate the solar declination angle through the obliquity of the ecliptic and the solar ecliptic longitude;

[0072] Based on the longitude and latitude coordinates of the unmanned vessel, combined with the hour angle and solar declination angle, the corresponding solar altitude angle is calculated, and the solar altitude angle sequence in the preset future period is obtained through iterative calculation;

[0073] Select the period in the solar altitude angle sequence when the solar altitude angle is greater than the preset altitude angle threshold as the optimal charging period of the day;

[0074] Collecting illumination data from the solar panel locations on the unmanned vessel. The illumination data includes the illumination sequence and cloud movement speed within a preset time period. This illumination data is used as input to a short-term prediction model to obtain the predicted illumination intensity within a preset future time period.

[0075] Obtain satellite images from radiation data, identify cloud pixels in the image through threshold segmentation, calculate the proportion of cloud area in the preset sea area, and mark it as cloud cover rate;

[0076] Obtain the cloud movement speed from the illumination data, use the cloud coverage and cloud movement speed as inputs to the cloud prediction model, and obtain the cloud coverage sequence within a preset future period;

[0077] If the predicted light intensity is greater than the preset light intensity threshold and the cloud coverage in the preset future time period is lower than the preset coverage threshold, the corresponding solar panel location is marked as a charging area, and the optimal charging time of the day is marked on the charging area.

[0078] The training methods for the cloud prediction model include:

[0079] A group B of cloud prediction data is collected in advance. The cloud prediction data includes cloud coverage and cloud movement speed, as well as the corresponding cloud coverage sequence within a preset future time period.

[0080] The cloud coverage and cloud movement speed are used as the input of the cloud prediction model, and the cloud coverage sequence within the corresponding preset future time period is used as the output of the cloud prediction model. The goal is to minimize the error between the output cloud coverage sequence within the corresponding preset future time period and the actual cloud coverage sequence within the corresponding preset future time period. The network parameters of the cloud prediction model are optimized through a nature-inspired optimization algorithm to obtain the network parameters that minimize the error between the cloud coverage sequence within the corresponding preset future time period output by the cloud prediction model and the actual cloud coverage sequence within the corresponding preset future time period. The cloud prediction model constructed with the corresponding network parameters is used as the trained cloud prediction model.

[0081] By monitoring light intensity in real time, the system dynamically generates a charging priority map, and in path planning, the charging benefit term in the cost function is used to guide the unmanned boat to prioritize high-light areas to recharge energy. When the battery charge is below the threshold, the data-driven system forces the path weight to be adjusted, and even if a detour is required, the path must be reached at the charging area. At the same time, it can combine the predicted drift trajectory with the predicted remaining battery life to ensure that the total path time is no longer than the predicted remaining battery life, avoiding mission interruption due to energy depletion. In addition, radiation data can also be linked with cloud cover predictions. When it is detected that the charging area is about to be blocked, path replanning is triggered in advance to achieve dynamic optimization of energy utilization.

[0082] Methods for obtaining the estimated remaining battery life include:

[0083] Obtain the current remaining power from the battery data, as well as the propeller power and mission equipment power consumption. Correct the propeller power according to the ocean current speed and the speed of the unmanned vessel to obtain the propeller power correction value. Calculate the total energy consumption based on the propeller power correction value and the mission equipment power consumption.

[0084] Obtain the output power, incident light power and panel area of the solar panel to obtain the conversion efficiency of the solar panel. Due to the limited deck area of the hull, in order to make full use of solar energy resources, high-wattage flexible photovoltaic solar panels are selected, and the size and specifications are customized according to the surface shape of the hull to achieve maximum curved laying and increase the solar energy collection area. For details, please refer to Figure 2 In addition to laying solar panels on the hull deck, it is also possible to consider installing some solar panels on the mast bracket to further expand the solar energy collection area; the lithium batteries that store the electricity converted by the solar panels will also be placed in a specially designed battery compartment at the bottom of the unmanned boat. The battery compartment adopts multiple protection measures such as sealing, waterproofing, fireproofing, and heat insulation to ensure that the batteries work in a safe and stable environment. At the same time, it can effectively reduce the impact of the battery pack weight on the center of gravity of the hull and improve the navigation stability of the hull.

[0085] The charge capacity corresponding to the current moment is calculated based on the light intensity and the conversion efficiency of the solar panel. The charge capacity at different moments is calculated based on the predicted light intensity and the conversion efficiency of the solar panel in the preset future period. The charge capacity at different moments in the preset future period is accumulated to obtain the charge capacity in the preset future period.

[0086] The remaining battery life is calculated based on the current remaining battery power, the charge level in the preset future period, and the total energy consumption.

[0087] By real-time monitoring of the battery's remaining charge, discharge rate, and historical energy consumption data, the remaining battery life is accurately predicted, providing key constraints for path planning: In global path search, the remaining battery life directly determines the upper limit of the total path time, forcing the algorithm to select the feasible path with the lowest energy consumption or the greatest charging benefit; in the dynamic adjustment phase, when the remaining battery life falls below the preset battery life safety threshold (such as 2 hours), a low-power mode or a loss reduction strategy of prioritizing the charging area is triggered, and the A* algorithm is used to replan the shortest charging path. In addition, the predicted remaining battery life is coordinated with the predicted drift trajectory. By calculating the ratio of the total path time (including drift offset) to the remaining battery life, the feasibility of the path is evaluated in real time, ensuring that the unmanned vessel can maintain energy balance despite environmental interference and avoiding mission interruption or device loss due to power exhaustion.

[0088] Methods for obtaining the global optimal path include:

[0089] Obtain charging areas, drift risk areas, range restricted areas, and preset optimization targets;

[0090] The drift risk zone is where the predicted ocean current speed exceeds the preset ocean current speed threshold, and the endurance restricted zone is where the total route time exceeds the remaining endurance time. The preset optimization goals include minimizing total energy consumption, maximizing charging benefits, and ensuring that the total route time is less than the remaining endurance time.

[0091] Calculate the Euclidean distance between the location data and the charging area; calculate the ratio of the electrical energy ultimately stored in the battery to the total energy of the incident light to obtain the energy conversion efficiency; calculate the product of the light intensity, the effective area of the solar panel, and the nominal efficiency of the solar panel to obtain the input power of the solar panel; calculate the product of the energy conversion efficiency, the input power of the solar panel, and the effective charging time to obtain the charging benefit; and weight the Euclidean distance, energy conversion efficiency, and charging benefit to calculate the cost function.

[0092] A heuristic function is constructed based on the weighted Euclidean distance and charging zone attractiveness. The weights of the Euclidean distance and charging zone attractiveness can be obtained through optimization using a nature-inspired optimization algorithm. The weight of the charging zone attractiveness is negative. The charging zone attractiveness is obtained by summing the attractiveness of all charging zones. The attractiveness is the ratio of the corresponding light intensity of the charging zone to the Euclidean distance. The closer to the charging zone, the lower the heuristic value. By minimizing the cost function and combining it with the heuristic function to expand nodes, the global optimal path is obtained.

[0093] Preset trigger conditions for dynamic path adjustment. When the trigger conditions are met, the global optimal path is replanned. The trigger conditions include: the deviation between the actual drifting trajectory and the predicted drifting trajectory is greater than the preset deviation threshold; the predicted remaining endurance time is lower than the preset endurance safety threshold;

[0094] Trigger conditions may also include sudden environmental changes (such as sudden storms (wind speed is greater than a preset wind speed threshold, wave height is greater than a preset wave height threshold), cloud coverage of the charging area (the rate of change of light intensity is higher than a preset change rate threshold, and the cloud coverage of the charging area is greater than a preset cloud coverage threshold), sudden changes in water temperature (the rate of change of water temperature is greater than a preset change threshold) and communication interruption (loss of connection with the control center exceeds a preset loss of connection duration), etc.).

[0095] The above steps utilize a multi-objective optimization strategy, using the predicted drift trajectory as an environmental constraint, the charging area as an energy gain point, and the remaining endurance time as a time window constraint. By combining the global optimal search of the A* algorithm with the local dynamic adjustment of the Dijkstra algorithm, a feasible path that balances shortest distance, low energy consumption, high charging benefits, and risk aversion is generated in real time. The A* algorithm's heuristic function can quickly find the initial optimal path; using the remaining endurance time as a hard constraint ensures the path's feasibility. This method significantly improves the path robustness, energy efficiency, and mission completion rate of unmanned vessels in complex marine environments.

[0096] Preset multiple power thresholds. When the current remaining battery power in the battery data falls below different thresholds, the corresponding loss reduction strategy is activated.

[0097] Methods for initiating corresponding loss reduction strategies include:

[0098] The first, second and third thresholds are preset. When the battery level is lower than the first threshold, the preset secondary tasks are suspended. When the battery level is lower than the second threshold, only the preset core tasks are retained. When the battery level is lower than the third threshold, the vehicle is forced to return home or enter sleep mode, and the return route planning prioritizes the charging area.

[0099] The above method divides the battery power into a safe zone (such as ≥50%), a warning zone (such as 30%-50%) and a critical zone (such as <30%) through the first threshold, the second threshold and the third threshold, and can dynamically adjust the path planning priority and control strategy; in the warning zone, the system increases the charging benefit weight and forces the path to tilt towards the high-light area; in the critical zone, it triggers the low-power mode (such as reducing the propeller power) and starts the Dijkstra algorithm to re-plan the shortest charging path. At the same time, the predicted remaining endurance time and the predicted drifting trajectory are combined to ensure that the total path time is ≤ the remaining endurance time, avoiding loss of control due to power exhaustion; it significantly improves the energy utilization and survivability of unmanned ships in long-duration missions, ensuring their continuous operation capabilities in complex environments.

[0100] Calculate path deviation and heading deviation based on position data and the global optimal path, use the path deviation, heading deviation, ocean current data, environmental data, battery data, and the global optimal path as inputs to the path tracking model to obtain the propulsion control parameters of the unmanned vessel; make real-time adjustments based on the propulsion control parameters of the unmanned vessel to achieve path tracking control; the propulsion control parameters of the unmanned vessel include the propeller thrust adjustment amount and the rudder deflection angle;

[0101] Methods for implementing path tracking control include:

[0102] Based on the position data of the unmanned vessel, the difference between the actual position and the nearest point of the preset path is calculated to obtain the path deviation; the difference between the current heading and the preset heading is calculated to obtain the heading deviation;

[0103] Path deviation, heading deviation, ocean current data, environmental data, battery data and the global optimal path are used as inputs of the path tracking model to obtain the unmanned ship propulsion control parameters; real-time adjustments are made based on the unmanned ship propulsion control parameters to achieve path tracking control.

[0104] The training methods for the path tracking model include:

[0105] F groups of path tracking data are collected in advance. The path tracking data include path deviation, heading deviation, ocean current data, environmental data, battery data, global optimal path, and corresponding unmanned vessel propulsion control parameters.

[0106] The path deviation, heading deviation, ocean current data, environmental data, battery data and the global optimal path are used as the input of the path tracking model, and the corresponding unmanned ship propeller control parameters are used as the output of the path tracking model. The goal is to minimize the error between the output corresponding unmanned ship propeller control parameters and the actual corresponding unmanned ship propeller control parameters. The network parameters of the path tracking model are optimized through the nature-inspired optimization algorithm to obtain the network parameters that minimize the error between the corresponding unmanned ship propeller control parameters output by the path tracking model and the actual corresponding unmanned ship propeller control parameters. The path tracking model constructed with the corresponding network parameters is used as the trained path tracking model.

[0107] The above steps compensate for path deviation (lateral error) and heading deviation (angular error) in real time, dynamically adjusting the propeller thrust and rudder angle through the path tracking model to reduce the deviation between the navigation trajectory and the planned path. They also use ocean current data to predict interference forces, feed-forward compensate for the impact of water currents on the hull, and reduce trajectory deviations caused by environmental disturbances. They can also optimize propulsion strategies based on the remaining battery power, reducing thrust to extend endurance when the battery is low, while maintaining course through fine-tuning the rudder angle. In complex scenarios (such as narrow river channels), navigation safety is ensured through control parameter constraints (such as maximum rudder angle and minimum thrust). This closed-loop control architecture achieves a seamless transition from path planning to physical execution, significantly improving the tracking accuracy and anti-interference capabilities of the unmanned vessel in dynamic marine environments.

[0108] Example 2

[0109] See also Figure 3 As shown, this embodiment provides a method for determining whether an unmanned vessel exceeds a preset boundary based on the position data of the unmanned vessel using a boundary judgment algorithm based on approximate convex hull decomposition, including the following steps:

[0110] Step 1: Obtain the coordinate sequence of the boundary vertices of the unmanned ship or the preset observation area and arrange them in a clockwise or counterclockwise order;

[0111] Step 2: Traverse all vertices and calculate the vector direction of the predecessor vertex and the successor vertex of each vertex;

[0112] Step 3: Calculate the difference between the vertex and the predecessor vertex, as well as the difference between the successor vertex and the vertex, and then calculate the cross product of the two differences. Get the sign of the cross product. If the sign is negative, the vertex is a concave point, otherwise it is a convex point. Record all concave points and generate a concave point list.

[0113] Step 4: Decompose the polygonal outline into two sub-polygons using the selected segmentation line; wherein the polygon is composed of a list of concave points of the unmanned ship or the preset observation area;

[0114] Step 5: Repeat steps 2 to 4 to perform concave point detection and segmentation on each sub-polygon until all sub-polygons are convex polygons.

[0115] Step 6: If the result of merging two adjacent sub-polygons is still a convex polygon, merge them, and give priority to merging the sub-polygons that share the longest common edge;

[0116] Step 7: Summarize the merged results to obtain a vertex set of the unmanned ship or the convex sub-polygon of the preset observation area;

[0117] Step 8: Obtain the boundary coordinates corresponding to the vertex set of the convex sub-polygon of the pre-calibrated observation area through GIS;

[0118] Step 9. Obtain the longitude and latitude coordinates corresponding to the vertex set of the convex polygon of the unmanned ship through GIS, and calculate the Euclidean distance between the longitude and latitude coordinates corresponding to the vertex set of the convex polygon of the unmanned ship and the boundary coordinates corresponding to the vertex set of the convex polygon of the observation area in real time. If the Euclidean distance exceeds the preset distance threshold, the path replanning is immediately triggered.

[0119] The methods for selecting the dividing line include:

[0120] For each concave point, connect it to other vertices and select a segmentation line that satisfies the following conditions:

[0121] Condition 1: The sub-polygon after segmentation contains the least number of concave points;

[0122] Condition 2: The splitting line does not intersect with other edges.

[0123] The above method steps can quickly determine whether the unmanned ship exceeds the preset boundary through convex hull containment judgment, triggering path replanning or thruster emergency steering; decomposing the irregular observation area into a convex hull set to reduce the geometric calculation complexity of boundary judgment, which is suitable for dynamically expanding ocean monitoring scenarios; providing accurate convex hull boundary constraints for the A* algorithm to ensure that the global optimal path is always within the safe area, avoiding the discrete errors of the traditional grid method; combining the predicted drift trajectory with ocean current data to predict the risk of crossing the boundary in the future period, adjust the path direction in advance, and achieve forward-looking safety control; when the charging area is on the convex hull boundary, the nearest charging path is quickly calculated through convex hull decomposition to balance the endurance requirements and safety constraints. The above method significantly improves the real-time and geometric adaptability of boundary judgment, and provides an efficient boundary management method for the safe and autonomous navigation of unmanned ships in complex waters.

[0124] Example 3

[0125] See also Figure 4 As shown, the unmanned ship adaptive path tracking control device described in this embodiment includes the following steps:

[0126] Boundary judgment module: collects the position data of the unmanned ship and uses the boundary judgment algorithm to determine whether the unmanned ship exceeds the preset boundary based on the position data of the unmanned ship. If it exceeds, it proceeds to the next step, otherwise it continues to judge;

[0127] Path planning module: collects and analyzes ocean current data, environmental data, and unmanned vessel data corresponding to the location data to obtain the predicted drift trajectory of the unmanned vessel; collects and analyzes radiation data corresponding to the location data to obtain the charging area; collects and analyzes battery data to obtain the predicted remaining battery life; constructs a path planning target based on the predicted drift trajectory, charging area, and predicted remaining battery life, and uses the path planning algorithm to perform dynamic path search to obtain the global optimal path;

[0128] Adaptive loss reduction module: preset multiple power thresholds. When the current remaining battery power in the battery data falls below different thresholds, the corresponding loss reduction strategy is activated;

[0129] Path tracking module: Calculates path deviation and heading deviation based on position data and the global optimal path, uses path deviation, heading deviation, ocean current data, environmental data, battery data, and the global optimal path as inputs to the path tracking model to obtain the propulsion control parameters of the unmanned vessel; performs real-time adjustments based on the propulsion control parameters of the unmanned vessel to achieve path tracking control.

[0130] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any modifications or substitutions that can be easily conceived by a person skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be based on the scope of protection of the claims.

[0131] Finally: The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. The unmanned ship adaptive path tracking control method is characterized by: The steps include: Collect the position data of the unmanned ship, and use the boundary judgment algorithm to determine whether the unmanned ship exceeds the preset boundary based on the position data of the unmanned ship. If it exceeds, proceed to the next step, otherwise continue to judge; Collect and analyze ocean current data, environmental data, and unmanned vessel data corresponding to the location data to obtain the predicted drift trajectory of the unmanned vessel; collect and analyze radiation data corresponding to the location data to obtain the charging area; collect and analyze battery data to obtain the predicted remaining flight time; Build a path planning target based on the predicted drift trajectory, charging area, and predicted remaining battery life, and use the path planning algorithm to dynamically search for paths to obtain the global optimal path; Methods for obtaining charging areas include: Obtain the UTC time corresponding to the collected unmanned vessel's position data, convert the UTC time to Julian day, calculate the Greenwich sidereal time based on the Julian day, and calculate the local sidereal time based on the Greenwich sidereal time; Calculate the mean ecliptic longitude and the correction term for the Earth's orbit eccentricity based on the Julian day, calculate the solar ecliptic longitude based on the mean ecliptic longitude and the correction term for the Earth's orbit eccentricity, calculate the solar right ascension based on the solar ecliptic longitude, and obtain the hour angle based on the local sidereal time and the solar right ascension; Calculate the solar declination angle through the obliquity of the ecliptic and the solar ecliptic longitude; Based on the longitude and latitude coordinates of the unmanned vessel, combined with the hour angle and solar declination angle, the corresponding solar altitude angle is calculated, and the solar altitude angle sequence in the preset future period is obtained through iterative calculation; Select the period in the solar altitude angle sequence when the solar altitude angle is greater than the preset altitude angle threshold as the optimal charging period of the day; Collecting illumination data from the solar panel locations on the unmanned vessel. The illumination data includes the illumination sequence and cloud movement speed within a preset time period. This illumination data is used as input to a short-term prediction model to obtain the predicted illumination intensity within a preset future time period. Obtain satellite images from radiation data, identify cloud pixels in the image through threshold segmentation, calculate the proportion of cloud area in the preset sea area, and mark it as cloud cover rate; Obtain the cloud movement speed from the illumination data, use the cloud coverage and cloud movement speed as inputs to the cloud prediction model, and obtain the cloud coverage sequence within a preset future period; If the predicted light intensity is greater than the preset light intensity threshold, and the cloud coverage rate in the preset future period is lower than the preset coverage rate threshold, the corresponding solar panel location is marked as a charging area, and the optimal charging time of the day is marked on the charging area; Preset multiple power thresholds. When the current remaining battery power in the battery data falls below different thresholds, the corresponding loss reduction strategy is activated. The path deviation and heading deviation are calculated based on the position data and the global optimal path. The path deviation, heading deviation, ocean current data, environmental data, battery data and the global optimal path are used as inputs of the path tracking model to obtain the unmanned ship propulsion control parameters; real-time adjustments are made based on the unmanned ship propulsion control parameters to achieve path tracking control.

2. The adaptive path tracking control method for an unmanned ship according to claim 1, characterized in that: The method for obtaining the predicted drift trajectory of the unmanned ship includes: According to the ocean current data, environmental data and unmanned ship data corresponding to the position data, based on Newton's laws of motion and fluid dynamics, combined with the ocean current data, environmental data and unmanned ship data corresponding to the position data, the motion equation of the unmanned ship under external force is established, the acceleration of the unmanned ship is extracted from the unmanned ship data, and the position and velocity of the unmanned ship in the next time step are calculated by combining the velocity-acceleration formula and the displacement-velocity formula; the position sequence within the preset trajectory period is obtained through iterative calculation, and the position sequence is used as the predicted drift trajectory of the unmanned ship.

3. The adaptive path tracking control method for an unmanned ship according to claim 1, characterized in that: The method for obtaining the predicted remaining battery life includes: Obtain the current remaining power from the battery data, as well as the propeller power and mission equipment power consumption. Correct the propeller power according to the ocean current speed and the speed of the unmanned vessel to obtain the propeller power correction value. Calculate the total energy consumption based on the propeller power correction value and the mission equipment power consumption. Obtain the output power, incident light power and panel area of the solar panel to obtain the conversion efficiency of the solar panel; The charge capacity corresponding to the current moment is calculated based on the light intensity and the conversion efficiency of the solar panel. The charge capacity at different moments is calculated based on the predicted light intensity and the conversion efficiency of the solar panel in the preset future period. The charge capacity at different moments in the preset future period is accumulated to obtain the charge capacity in the preset future period. The remaining battery life is calculated based on the current remaining battery power, the charge level in the preset future period, and the total energy consumption.

4. The adaptive path tracking control method for an unmanned vessel according to claim 1, characterized in that: The method for obtaining the global optimal path includes: Obtain charging areas, drift risk areas, range restricted areas, and preset optimization targets; The drift risk zone is where the predicted ocean current speed exceeds the preset ocean current speed threshold, and the endurance restricted zone is where the total route time exceeds the remaining endurance time. The preset optimization goals include minimizing total energy consumption, maximizing charging benefits, and ensuring that the total route time is less than the remaining endurance time. Calculate the Euclidean distance between the location data and the charging area; calculate the ratio of the electrical energy ultimately stored in the battery to the total energy of the incident light to obtain the energy conversion efficiency; calculate the product of the light intensity, the effective area of the solar panel, and the nominal efficiency of the solar panel to obtain the input power of the solar panel; calculate the product of the energy conversion efficiency, the input power of the solar panel, and the effective charging time to obtain the charging benefit; and weight the Euclidean distance, energy conversion efficiency, and charging benefit to calculate the cost function. A heuristic function is constructed based on the Euclidean distance and the weighted attractiveness of the charging area, where the weight of the attractiveness of the charging area is a negative value. By minimizing the cost function and combining the heuristic function to expand the nodes, the global optimal path is obtained.

5. The adaptive path tracking control method for an unmanned ship according to claim 4, characterized in that: The charging area attraction is obtained by summing the attractions of all charging areas, and the attraction is the ratio of the light intensity corresponding to the charging area to the Euclidean distance.

6. The unmanned vessel adaptive path tracking control method according to claim 4, characterized in that: In the method for obtaining the global optimal path, trigger conditions for dynamic path adjustment are also preset. When the trigger conditions are met, the global optimal path is replanned; the trigger conditions include: the deviation value between the actual drifting trajectory and the predicted drifting trajectory is greater than a preset deviation threshold; the predicted remaining endurance time is lower than a preset endurance safety threshold.

7. The unmanned vessel adaptive path tracking control method according to claim 1, characterized in that: The method for determining whether the unmanned ship exceeds the preset boundary based on the position data of the unmanned ship based on the boundary judgment algorithm includes: Preset the boundaries of the observation area and pre-calibrate the boundary coordinates of the observation area through GIS; The current longitude and latitude coordinates of the unmanned ship are obtained through GIS, and the Euclidean distance between the current longitude and latitude coordinates of the unmanned ship and the boundary coordinates is calculated in real time. If the Euclidean distance exceeds the preset distance threshold, the path replanning is triggered immediately.

8. The adaptive path tracking control method for an unmanned vessel according to claim 1, wherein: The method of starting the corresponding loss reduction strategy includes: The first, second, and third thresholds are preset. When the battery level drops below the first threshold, the preset secondary tasks are suspended. When the battery level drops below the second threshold, only the preset core tasks are retained. When the battery level drops below the third threshold, the aircraft is forced to return home or enter sleep mode, with the return route planned to prioritize the charging area. Path deviation, heading deviation, ocean current data, environmental data, battery data and the global optimal path are used as inputs of the path tracking model to obtain the unmanned ship propulsion control parameters; real-time adjustments are made based on the unmanned ship propulsion control parameters to achieve path tracking control.

9. The adaptive path tracking control method for an unmanned vessel according to claim 1, characterized in that: The method for implementing path tracking control comprises: Based on the position data of the unmanned vessel, the difference between the actual position and the nearest point of the preset path is calculated to obtain the path deviation; the difference between the current heading and the preset heading is calculated to obtain the heading deviation; Path deviation, heading deviation, ocean current data, environmental data, battery data and the global optimal path are used as inputs of the path tracking model to obtain the unmanned ship propulsion control parameters; real-time adjustments are made based on the unmanned ship propulsion control parameters to achieve path tracking control.

10. An unmanned vessel adaptive path tracking control device, implementing the unmanned vessel adaptive path tracking control method according to any one of claims 1 to 9, characterized in that: include: Boundary judgment module: collects the position data of the unmanned ship and uses the boundary judgment algorithm to determine whether the unmanned ship exceeds the preset boundary based on the position data of the unmanned ship. If it exceeds, it proceeds to the next step, otherwise it continues to judge; Path planning module: collects and analyzes ocean current data, environmental data, and unmanned vessel data corresponding to the location data to obtain the predicted drift trajectory of the unmanned vessel; collects and analyzes radiation data corresponding to the location data to obtain the charging area; collects and analyzes battery data to obtain the predicted remaining flight time; Build a path planning target based on the predicted drift trajectory, charging area, and predicted remaining battery life, and use the path planning algorithm to dynamically search for paths to obtain the global optimal path; Adaptive loss reduction module: preset multiple power thresholds. When the current remaining battery power in the battery data falls below different thresholds, the corresponding loss reduction strategy is activated; Path tracking module: Calculates path deviation and heading deviation based on position data and the global optimal path, uses path deviation, heading deviation, ocean current data, environmental data, battery data, and the global optimal path as inputs to the path tracking model to obtain the propulsion control parameters of the unmanned vessel; performs real-time adjustments based on the propulsion control parameters of the unmanned vessel to achieve path tracking control.

Images