Unmanned ship formation path planning method and system based on cloud control

Abstract

The invention relates to the field of unmanned ship technology control, in particular to an unmanned ship formation path planning method and system based on cloud control. Comprising the following steps: acquiring initial position information and target position information of each unmanned ship in an unmanned ship formation and environment information in a navigation area through a cloud platform; constructing a digital map of the navigation area through the cloud platform; planning a plurality of candidate paths for each unmanned ship through a digital map; calculating a cost function value of each candidate path through a preset cost model, and screening out an initial optimal planning path; the position information of each unmanned ship and the environment information in the navigation area are received in real time through the cloud platform, and the optimal planning path is updated. According to the invention, the position information of all unmanned ships in the formation and the environment data of the navigation area are acquired in a centralized manner through the cloud platform, a unified digital map is constructed and updated in real time, and the limitation of local sensing of a single ship in distributed control is broken through.

Description

Technical Field

[0001] The present invention relates to the field of unmanned ship technology control, and in particular to a cloud-controlled unmanned ship formation path planning method and system. Background Art

[0002] Unmanned vessels are surface robots that can navigate waters according to pre-set missions without remote control, relying on precise satellite positioning and their own sensors. These "surface robots" integrate multiple technologies, including ship control, communications, automated control, remote monitoring, and networked systems, enabling autonomous navigation, intelligent obstacle avoidance, long-distance communication, real-time video transmission, and networked monitoring. As a new technological tool, unmanned vessels have been widely used in fields such as oceanographic surveying and maritime defense.

[0003] Among existing unmanned vessel formation path planning methods, some schemes adopt a decentralized control strategy, relying on individual vessels to locally store route data and autonomously adjust speed to maintain formation. However, this model has significant limitations: due to the lack of globally unified environmental perception and path optimization capabilities, individual vessels can only make decisions based on preset routes and local information, making it difficult to cope with dynamic environmental changes such as the sudden appearance of obstacles and sudden changes in water speed and direction within the navigation area. At the same time, under decentralized control, the path planning of each vessel is relatively independent, making it difficult to coordinate and optimize the safety and efficiency of the path from a global perspective of the formation. This can easily lead to local path redundancy, increased overall energy consumption of the formation, and even collision risks due to conflicting trajectories of each vessel, seriously affecting the reliability of the unmanned vessel formation's mission execution in complex dynamic environments.

[0004] Therefore, it is necessary to design a cloud-controlled unmanned ship formation path planning method and system. Summary of the Invention

[0005] In response to the technical deficiencies in the background technology, the present invention proposes a cloud-controlled unmanned vessel formation path planning method and system, which solves the above technical problems and meets practical needs. The specific technical solution is as follows:

[0006] The cloud-controlled unmanned vessel formation path planning method includes the following steps:

[0007] Obtain the initial position information, target position information and environmental information of each unmanned ship in the unmanned ship formation through the cloud platform;

[0008] Based on environmental information, a digital map of the navigation area is constructed through the cloud platform;

[0009] Based on the initial position information and target position information of each unmanned vessel, multiple candidate paths are planned for each unmanned vessel through a digital map;

[0010] Based on multiple candidate paths, the cost function value of each candidate path is calculated through a preset cost model, and the initial optimal planning path is screened out based on the cost function values ​​of the multiple candidate paths;

[0011] During the navigation process of each unmanned vessel according to the initial optimal planned path, the cloud platform receives its own position information and environmental information in the navigation area in real time, updates the digital map based on the environmental information, and updates the optimal planned path based on the updated digital map and its own position information.

[0012] Furthermore, the environmental information includes obstacle distribution information, water flow speed and direction information, and the digital map divides the navigation area into multiple grid cells, each grid cell corresponding to a unique coordinate. The digital map of the navigation area constructed through the cloud platform specifically includes:

[0013] Based on the obstacle distribution information, the grid cells containing obstacles are marked as impassable cells through the cloud platform;

[0014] Based on the water flow speed and direction information, the corresponding water flow impact parameters are assigned to each grid unit through the cloud platform.

[0015] Furthermore, the planning of multiple candidate paths for each unmanned vessel using the digital map specifically includes:

[0016] A heuristic search algorithm is used, with the grid cell corresponding to the initial position of the unmanned vessel as the starting point and the grid cell corresponding to the target position as the end point, to search and generate multiple paths leading to the end point on the digital map as candidate paths;

[0017] During the search process, ensure that the candidate path does not pass through the grid cells marked as impassable cells.

[0018] Furthermore, the cost function value of each candidate path is calculated using a preset cost model and specifically includes:

[0019] Calculate the number of grid cells that the candidate path passes through and determine the path length cost;

[0020] Calculate the distance between the center of each grid cell on the candidate path and the nearest obstacle, accumulate the inverse of the distance, and determine the distance cost to the obstacle;

[0021] Based on the water flow impact parameters and the angle between the unmanned boat's navigation direction and the water flow direction, the water flow impact component in each grid cell is calculated. The water flow impact components in all grid cells on the candidate path are accumulated to determine the water flow impact cost.

[0022] The path length cost, obstacle distance cost and water flow impact cost are weighted and summed according to the preset weight coefficient to obtain the cost function value of the candidate path.

[0023] Furthermore, the real-time receiving of the position information of each unmanned vessel and the environmental information within the navigation area through the cloud platform specifically includes:

[0024] The unmanned boat obtains its own location information through its own positioning equipment and sends its location information to the cloud platform through the wireless communication module;

[0025] The unmanned boat uses its own sensors to detect new obstacles in the navigation area and changes in water speed and direction, and sends the obstacle information and change information to the cloud platform through the wireless communication module.

[0026] Furthermore, each unmanned vessel, while navigating according to the initial optimal planned path, further comprises the following steps:

[0027] Based on the initial optimal planning path of each unmanned ship in the unmanned ship formation, the predicted trajectories of any two unmanned ships in the future are calculated and the predicted trajectories of the two unmanned ships are judged;

[0028] If there is a possibility of intersection between the predicted trajectories, the initial optimal planned path of one of the unmanned ships is adjusted according to the priority. Based on the adjusted optimal planned path, the adjusted unmanned ship is recalculated to see whether there is a possibility of intersection between the predicted trajectories of other unmanned ships in the unmanned ship formation. If there is a possibility of intersection, the optimal planned path of the unmanned ship is readjusted until there is no possibility of intersection between the predicted trajectories of the adjusted unmanned ship and other unmanned ships in the unmanned ship formation. If not, other pairs of unmanned ships are continued to be checked until there is no possibility of intersection between the predicted trajectories of all unmanned ships.

[0029] Furthermore, the determination of the predicted trajectories of the two unmanned ships specifically includes:

[0030] Obtain the initial planned path, maximum navigation speed, minimum turning radius and current position information of any two unmanned ships in the unmanned ship formation, and calculate the predicted position coordinates of the two unmanned ships at several future time nodes based on the preset time interval;

[0031] Based on the predicted position coordinates of the two unmanned ships at several future time points, the distance between the predicted positions of the two unmanned ships at the same time point is calculated. If the distance is less than the preset safety distance threshold and the predicted trajectories of the two unmanned ships before and after the time point show a trend of gradually approaching, it is determined that there is a possibility that the predicted trajectories of the two unmanned ships intersect;

[0032] The safety distance threshold is dynamically adjusted according to the length, width and navigation speed of the unmanned ship.

[0033] Furthermore, adjusting the initial optimal planning path of one of the unmanned ships according to the priority specifically includes:

[0034] Determine the unmanned vessel that needs to adjust the optimal planned path based on mission priority, current load status, and remaining distance to the target location;

[0035] Reselect a path from multiple candidate paths whose cost function value is within a preset range and can avoid collision risks;

[0036] If there is no candidate path whose cost function value differs from the original path within the preset range, the original optimal planned path is locally optimized, and turning points are added near the predicted collision area during local optimization;

[0037] Among them, the setting of the turning point must meet the minimum turning radius constraint of the unmanned ship, and the path after turning must be staggered in time and space with the predicted trajectory of another unmanned ship.

[0038] Furthermore, updating the digital map using environmental information specifically includes:

[0039] Based on the newly detected obstacle information, the traffic status of the corresponding grid cell in the digital map is updated, and the grid cell where the new obstacle is located is marked as an impassable cell;

[0040] Update the water flow impact parameters of each grid cell in the digital map based on the new water flow speed and direction information;

[0041] Taking the grid cell corresponding to the current position of the unmanned ship as the new starting point and the grid cell corresponding to the target position as the end point, the heuristic search algorithm is re-adopted to generate multiple candidate paths and the optimal planning path is re-screened.

[0042] An unmanned vessel formation path planning system based on cloud control, comprising a processor and a memory for storing instructions executable by the processor;

[0043] The processor is configured to execute the instructions to implement the steps of the cloud-controlled unmanned ship formation path planning method as described above.

[0044] Compared with the existing technology, the cloud-controlled unmanned vessel formation path planning method and system provided by the present invention has the following beneficial effects:

[0045] This method uses a cloud platform to centrally obtain the location information of all unmanned vessels in a fleet and environmental data of the navigation area, constructing and updating a unified digital map in real time. This overcomes the limitations of decentralized control, where "local perception" is limited to individual vessels. Dynamically generating optimal paths based on the global digital map enables rapid response to sudden environmental changes, resolving the issue of decentralized approaches that rely on pre-set paths and struggle to adapt to dynamic environments.

[0046] This invention uses a pre-set cost model to comprehensively consider multiple factors, including path length, distance to obstacles, and the impact of water currents, to identify the globally optimal path. Compared to decentralized approaches that focus solely on maintaining formation through speed adjustments, this invention achieves a balance between safety and efficiency during the path planning phase, reducing ineffective navigation costs.

[0047] This invention uses a cloud platform to predict the future trajectory of each unmanned vessel in real time and adjust the global priority of potential intersecting trajectories, ensuring that vessels within the formation sail in staggered time and space. This mechanism is superior to the decentralized collision avoidance approach that relies on fine-tuning the speed of individual vessels. It can fundamentally avoid trajectory conflicts and improve the safety of formation coordination. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] Figure 1 The figure is a flow chart of the path planning method for the unmanned ship formation based on cloud control in the present invention. DETAILED DESCRIPTION

[0049] In the description of the present invention, it should be understood that the terms "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "center," and "inner" are used to indicate positions or locations based on those shown in the accompanying drawings. These terms are intended solely to facilitate the description of the present invention and simplify the description. They are not intended to indicate or imply that the devices or components referred to must have a specific orientation, be constructed, or operate in a specific orientation. Therefore, they should not be construed as limiting the present invention. Furthermore, the terms "first," "second," and the like are used for descriptive purposes only and should not be construed to indicate or imply relative importance or implicitly specify the number of the technical features referred to. Thus, features defined as "first," "second," and the like may explicitly or implicitly include one or more of such features. In the description of the present invention, it should be noted that, unless otherwise expressly specified or limited, the terms "mounted," "connected," and "connected" should be understood broadly. For example, they can refer to fixed connections, removable connections, or integral connections; they can refer to direct connections, indirect connections through an intermediary, or internal connections between two components. Those skilled in the art will understand the specific meanings of the above terms in the present invention based on specific circumstances.

[0050] The following describes the implementation of the present invention in conjunction with the accompanying drawings and relevant embodiments. The implementation of the present invention is not limited to the following embodiments, and the present invention relates to the relevant necessary components in this technical field, which should be regarded as the common knowledge in this technical field and can be known and mastered by technical personnel in this technical field.

[0051] See Figure 1The present invention provides a path planning method for an unmanned ship formation based on cloud control, comprising the following steps:

[0052] Step S100: obtaining the initial position information, target position information and environmental information within the navigation area of ​​each unmanned ship in the unmanned ship formation through the cloud platform.

[0053] The cloud platform is a distributed computing and storage system based on cloud computing technology. Its functions include data fusion, algorithm calculation, and command issuance. It can achieve collaborative operation of multiple terminal devices through the Internet of Things protocol. The initial position information is the geographic coordinate data of the unmanned vessel at the start of the mission, usually obtained through a satellite positioning system or an inertial navigation system. The target position information is the coordinates of the destination that each unmanned vessel needs to reach, which is pre-set or dynamically adjusted by the mission planning system. Environmental information includes the distribution of dynamic and static obstacles in the navigation area, the speed and direction of water flow, and meteorological data, which are collected in real time through sensor networks and external data interfaces.

[0054] This is achieved by establishing a communication link to receive and integrate heterogeneous data from multiple sources. Specifically, the unmanned vessel uploads real-time coordinates, the cloud platform receives target location instructions, and calls the environmental monitoring system to obtain real-time environmental parameters. This operation provides a global data foundation for subsequent path planning, overcoming the limitations of a single vessel's local perception and ensuring the integrity of environmental data.

[0055] Step S200: constructing a digital map of the navigation area through a cloud platform based on environmental information.

[0056] Among them, the digital map is a virtual space that digitally models the physical space, which contains multi-dimensional data of geographic coordinate system, obstacle contours, and hydrological characteristics, and its form is a raster map.

[0057] In step S300 , a plurality of candidate paths are planned for each unmanned vessel using a digital map based on the initial position information and target position information of each unmanned vessel.

[0058] Candidate paths are multiple feasible routes from the starting point to the end point, generated using a heuristic search algorithm. Digital map applications can use obstacles and hydrological data as constraints for path planning, such as setting no-fly zones or water flow weights.

[0059] This is achieved through path search and cost setting, specifically determining the path search range based on the initial and target locations, setting movement costs for path nodes, and using an algorithm to traverse feasible paths and record multiple routes that meet the constraints. This operation provides diverse path selection and provides alternative solutions for subsequent optimization.

[0060] Step S400 : Based on multiple candidate paths, a cost function value of each candidate path is calculated using a preset cost model, and an initial optimal planned path is screened out based on the cost function values ​​of the multiple candidate paths.

[0061] The cost model is a mathematical function that integrates path length, safety margin, energy consumption, and mission time. For example, the safety cost can be defined as the inverse of the minimum distance between the path and the obstacle, while the energy cost can be calculated by combining water flow direction and the ship's hull dynamics model. The cost function value is a comprehensive evaluation value obtained through weighted summation or product operations, with lower values ​​representing better paths.

[0062] Specifically, this involves defining a weighted distribution rule for the cost function, calculating the total length of each candidate path, assessing the minimum distance between each point on the path and obstacles, simulating the navigation time of the path under current conditions, and then using a weighted formula to determine the overall cost. This process can achieve a quantitative balance between safety and efficiency, for example, prioritizing paths that avoid strong currents to reduce energy consumption.

[0063] In step S500, while each unmanned vessel is navigating according to the initial optimal planned path, it receives its own position information and environmental information within the navigation area in real time through the cloud platform, updates the digital map based on the environmental information, and updates the optimal planned path based on the updated digital map and its own position information.

[0064] Real-time data reception is achieved through 5G / satellite communication technology, ensuring that the data update frequency matches the speed of the unmanned vessel. The digital map update mechanism is implemented through incremental updates, for example, only updating map data in the area where the obstacle moves.

[0065] This is achieved through closed-loop control and predictive algorithms, which involve continuously collecting data, detecting environmental changes, triggering local map updates, re-executing path planning, and issuing a new path. This operation enhances the ability to adapt to dynamic environments, such as quickly adjusting the path to maintain formation during sudden changes in water flow, while also avoiding global path redundancy caused by local obstacle avoidance.

[0066] This method integrates fleet-wide data through a cloud platform to eliminate the limitations of local decision-making by individual vessels. It constructs a real-time, updated global environmental model to support dynamic response, generates multiple candidate paths, and quantitatively selects the optimal path based on a cost model. Furthermore, it dynamically adjusts the path through closed-loop control. This method achieves the technical benefits of globally optimizing navigation efficiency and safety, rapidly responding to environmental changes, and reducing navigation redundancy. Through centralized computing and real-time feedback, this method addresses the shortcomings of traditional decentralized control in terms of dynamic environmental adaptability, global path optimization, and collaborative safety, providing reliable technical support for the execution of unmanned vessel formation missions in complex waters.

[0067] In one embodiment of the present invention, the environmental information includes obstacle distribution information, water flow speed and direction information, and the digital map divides the navigation area into multiple grid cells, each grid cell corresponding to a unique coordinate. The construction of the digital map of the navigation area through the cloud platform specifically includes: based on the obstacle distribution information, marking the grid cells containing obstacles as impassable cells through the cloud platform; and based on the water flow speed and direction information, assigning corresponding water flow impact parameters to each grid cell through the cloud platform.

[0068] Among them, the grid unit is a regular two-dimensional or three-dimensional space unit that divides the navigation area on the digital map. Each unit has a unique coordinate identifier and can be divided in a uniform grid or adaptive grid. Its function is to discretize the continuous space to facilitate the rapid retrieval and calculation of the path planning algorithm. The impassable unit is a grid unit occupied by obstacles or that poses a navigation risk. After obtaining the obstacle distribution information through sensors or external data sources, the grid covered by the obstacle is determined and marked through a coordinate matching algorithm. For example, if the coordinates (x, y) of a grid center point are detected by radar as a fixed obstacle, the grid and its adjacent areas may be marked as impassable as a whole. The water flow influencing parameter is the speed and direction data of the water flow in each grid unit, usually expressed in vector form. Its purpose is to quantify the resistance or assistance effect of the water flow on the navigation of the unmanned ship. For example, sailing downstream can reduce energy consumption, while sailing upstream requires increased power output.

[0069] The cloud platform first divides the navigation area into grid cells at a preset resolution, assigning unique coordinates to each cell. It then uses a coordinate mapping algorithm to determine impassable cells based on obstacle distribution information. This process can be achieved, for example, by matching obstacle locations with grid coordinates and marking no-fly zones, thereby directly eliminating collision-risk paths from path planning. For water flow parameters, the system spatially interpolates or distributes the collected water flow velocity and direction data according to grid locations. For example, bilinear interpolation is used to map continuous water flow field data to discrete grids. This process ensures the spatiotemporal consistency of water flow parameters. For example, higher-resolution grids are used in areas where the water flow direction changes suddenly to improve data accuracy, thereby providing accurate input for dynamic environmental impacts in subsequent cost models.

[0070] This invention achieves the technical effects of improving the efficiency of path planning algorithms, enhancing adaptability to dynamic environments, and optimizing the accuracy of energy consumption estimates by dividing the navigation area into grid cells with unique coordinates to structure environmental information, marking impassable cells to eliminate collision-risk paths, and assigning water flow influencing parameters to quantify the impact of the dynamic environment on navigation energy consumption. Specifically, the structured digital map enables the path planning algorithm to efficiently retrieve obstacle distribution and water flow characteristics, the local parameter update mechanism reduces computing resource consumption, and the heuristic search algorithm in discrete space can quickly converge on the path, ultimately enhancing environmental perception accuracy and dynamic response speed. For example, when the water flow suddenly changes, the path can be quickly adjusted through local grid parameter updates rather than relying on full area recalculation.

[0071] In one embodiment of the present invention, planning multiple candidate paths for each unmanned vessel using a digital map specifically includes: using a heuristic search algorithm, with the grid cell corresponding to the initial position of the unmanned vessel as the starting point and the grid cell corresponding to the target position as the end point, searching and generating multiple paths leading to the end point on the digital map as candidate paths; during the search process, ensuring that the candidate paths do not pass through grid cells marked as impassable cells.

[0072] Among them, the heuristic search algorithm is a path search method that evaluates the priority of node expansion through a heuristic function. Its core is to balance the efficiency and optimality of path exploration. Exemplarily, it includes the A-star algorithm, the DijkstrA-star algorithm or the RRT algorithm. The heuristic function is a mathematical model that estimates the remaining cost from the node to the end point. Exemplarily, the straight-line distance is calculated using Manhattan distance or Euclidean distance, and a comprehensive ranking is performed in combination with the actual cost parameters of the path length or energy consumption. The exclusion of impassable units is a path planning strategy that filters specific grid units through preset rules. Exemplarily, when the adjacent grid is marked as an obstacle area or a no-fly zone, the path expansion in that direction is directly terminated.

[0073] This can be achieved through the following steps: First, grid cells on a digital map are used as search nodes, with impassable cells pre-set to an inaccessible state. Second, weight parameters are defined for the heuristic function and the actual cost function. For example, the straight-line distance between the endpoint and the current node is used as the heuristic function value, combined with the energy consumption of movement as the actual cost. Third, a priority queue is used to expand path branches, prioritizing paths with the lowest total cost. Paths are checked in real time to see if they contain impassable cells, discarding those that do. Finally, multiple feasible candidate paths are generated by adjusting the heuristic function weights or introducing a path diversity parameter. For example, in the A-star algorithm, the conservatism of the heuristic function can be relaxed or random perturbations can be added to prevent paths from converging in a single direction. This ensures that paths avoid obstacles, improves search efficiency, and provides diverse path options for subsequent screening.

[0074] This invention combines a heuristic search algorithm with the structured data foundation of a gridded digital map, utilizing a comprehensive evaluation of heuristic functions and actual costs to achieve efficient path exploration. It also ensures path safety by excluding impassable elements and generates multiple candidate paths based on parameter adjustments to accommodate dynamic environmental changes. This approach can rapidly generate feasible paths in scenarios with complex obstacle distribution or sudden changes in water flow. Through the synergy of algorithms and data structures, this solution achieves efficient, secure, and adaptable path planning. For example, it can rapidly respond to changes in obstacles in a real-time digital map, or reduce energy consumption by adjusting the heuristic function to prioritize downstream paths.

[0075] In one embodiment of the present invention, the step of calculating the cost function value of each candidate path using a preset cost model specifically includes:

[0076] Calculate the number of grid cells that the candidate path passes through and determine the path length cost;

[0077] Calculate the distance between the center of each grid cell on the candidate path and the nearest obstacle, accumulate the inverse of the distance, and determine the distance cost to the obstacle;

[0078] Based on the water flow impact parameters and the angle between the unmanned boat's navigation direction and the water flow direction, the water flow impact component in each grid cell is calculated. The water flow impact components in all grid cells on the candidate path are accumulated to determine the water flow impact cost.

[0079] The path length cost, obstacle distance cost and water flow impact cost are weighted and summed according to the preset weight coefficient to obtain the cost function value of the candidate path.

[0080] The path length cost is a numerical indicator that measures the total length of a path by the number of grid cells or actual distance passed by the candidate path. It can be converted using the grid resolution of the digital map. For example, if each grid cell corresponds to an actual area of ​​10m×10m, the path length cost can directly use the number of grid cells or actual distance as the base value. Its purpose is to prioritize routes with shorter paths to reduce navigation time or energy consumption. In one specific embodiment, the calculation of the path length cost includes traversing each grid cell on the candidate path, counting the total number or converting the actual distance using a preset grid side length to form a basic value for the path length.

[0081] The obstacle distance cost is calculated by summing the reciprocal distances from each grid center point on the path to the nearest obstacle, quantifying the proximity of the path to the obstacle. For example, if a grid center is 5 meters away from an obstacle, its reciprocal value is 0.2; a grid center 10 meters away has a reciprocal value of 0.1. This design mathematically transforms "maximizing safety distance" into minimizing the cost function. In one specific embodiment, the obstacle distance cost is calculated by searching the digital map for the nearest obstacle distance for each grid center point on the path, calculating the reciprocal values, and summing them to form a comprehensive safety cost.

[0082] The water flow impact cost is calculated by quantifying the degree of directional matching using the cosine function of the angle between the water flow velocity vector and the unmanned vessel's navigation direction, and then adding up the comprehensive components calculated using the water flow velocity. For example, when the angle between the water flow direction and the navigation direction is 0°, the water flow impact component is positive; when the angle is 180°, the component is negative. In one specific embodiment, the calculation of the water flow impact cost involves extracting the water flow velocity vector and the navigation direction vector for each grid cell on the path, calculating the angle θ, substituting the calculated component values ​​into a formula, and summing the components for all grid cells.

[0083] Weighted summation involves linearly combining the path length cost, obstacle distance cost, and water flow impact cost by multiplying each by a preset weight coefficient. For example, the total cost can be expressed as: Total Cost = Weight 1 × Path Length Cost + Weight 2 × Obstacle Distance Cost + Weight 3 × Water Flow Impact Cost. The weight coefficients can be adjusted based on mission requirements. For example, in a safety-first scenario, the obstacle distance cost could be weighted 0.4, while the path length cost and water flow impact cost each account for 0.3. This operation achieves a multi-objective quantitative balance, enabling path planning to simultaneously meet requirements for range, safety, and energy efficiency.

[0084] This invention optimizes navigation efficiency by calculating the cost of path length, strengthens safety constraints through the inverse design of obstacle distance, improves the depth of energy consumption optimization through refined modeling of the angle between water flow direction and navigation direction, and achieves multi-objective dynamic balance through weight coefficients. This method can achieve the coordinated optimization of path length, safety distance, and energy consumption in complex water environments, while also having the technical effect of dynamic environmental responsiveness and improved configurability. This solution significantly improves the scientific nature and adaptability of path planning through the spatial benchmark of gridded digital maps, the screening of candidate paths generated by heuristic search, and the recalculation of costs triggered by real-time environmental parameter updates. For example, when the water flow suddenly changes or the obstacle distribution changes, the system can quickly generate an optimal path that takes into account both downstream navigation and obstacle avoidance, thereby improving the execution efficiency and reliability of formation tasks.

[0085] In one embodiment of the present invention, the real-time receiving of the self-position information of each unmanned vessel and the environmental information within the navigation area through the cloud platform specifically includes:

[0086] The unmanned boat obtains its own location information through its own positioning equipment and sends its location information to the cloud platform through the wireless communication module;

[0087] The unmanned boat uses its own sensors to detect new obstacles in the navigation area and changes in water speed and direction, and sends the obstacle information and change information to the cloud platform through the wireless communication module.

[0088] Among them, the positioning device is a coordinate positioning device used to determine the real-time geographic coordinates of the unmanned ship, including a satellite positioning module or an inertial navigation system. Exemplarily, the satellite positioning module can be a device such as GPS, Beidou, etc. that obtains latitude and longitude coordinates by receiving satellite signals, with an accuracy of up to meter level; the inertial navigation system measures displacement and heading through accelerometers and gyroscopes, and is suitable for areas where satellite signals are blocked. The wireless communication module is a communication device that supports two-way data transmission, and its technology types include but are not limited to 5G, LoRa, satellite communication or cellular network. In a specific embodiment, the module achieves stable data transmission through low-latency transmission and high-bandwidth characteristics, and has anti-interference capabilities to cope with complex electromagnetic environments.

[0089] Sensors are devices used to detect changes in the navigation environment. They include radar, Doppler current profilers, sonar, and meteorological sensors. For example, radar detects the distance and direction of obstacles, Doppler current profilers measure water speed and direction, sonar detects underwater obstacles, and meteorological sensors monitor parameters such as wind speed and wave height. Sensor data undergoes preprocessing steps, such as filtering and denoising, or coordinate system conversion, to ensure correlation with the grid cells of the digital map.

[0090] Position information collection and transmission is achieved through the following method: The unmanned vessel's positioning device continuously receives satellite signals or calculates displacement through inertial navigation, generating position coordinates at preset intervals. The wireless communication module then packages the coordinate data into a standardized protocol, adds a timestamp to ensure data synchronization, and transmits it to the cloud platform. This process improves the real-time nature of data collection, and the redundant design of satellite positioning and inertial navigation reduces the risk of data loss due to single device failure.

[0091] The present invention uses the positioning equipment and wireless communication module on the unmanned vessel to achieve real-time collection and transmission of location information, acquires environmental change information through the collaborative detection and data fusion technology of multiple types of sensors, and optimizes bandwidth utilization efficiency through the intelligent scheduling strategy of the wireless communication module. The following technical effects can be achieved: the cloud platform can quickly respond to local environmental changes, dynamically update the grid cell status of the digital map, and combine the prediction algorithm to predict the trajectory of the unmanned vessel and trigger path replanning, thereby improving the system's adaptability to complex dynamic environments. For example, in dense waterways, high-precision digital maps can be constructed through the fusion of multi-vessel sensor data, or the reliability of the formation mission can be maintained through real-time path replanning when sudden water flow changes occur, ultimately achieving a comprehensive improvement in the real-time performance, accuracy, and anti-interference capability of the formation path planning.

[0092] In one embodiment of the present invention, when each unmanned ship navigates according to the initial optimal planned path, the following steps are also included: based on the initial optimal planned path of each unmanned ship in the unmanned ship formation, the predicted trajectories of any two unmanned ships in the future period are calculated, and the predicted trajectories of the two unmanned ships are judged; if there is a possibility of intersection between the predicted trajectories, the initial optimal planned path of one of the unmanned ships is adjusted according to the priority, and based on the adjusted optimal planned path, whether there is a possibility of intersection between the adjusted unmanned ship and the predicted trajectories of other unmanned ships in the unmanned ship formation is recalculated; if there is a possibility of intersection, the optimal planned path of the unmanned ship is readjusted until there is no possibility of intersection between the adjusted unmanned ship and the predicted trajectories of other unmanned ships in the unmanned ship formation; if not, other pairs of unmanned ships are continued to be checked until the predicted trajectories of all unmanned ships have no possibility of intersection.

[0093] Among them, the predicted trajectory is the result of mathematical deduction of the track in the future time window based on the unmanned ship motion model, and is generated through a kinematic model or a dynamic model. For example, the kinematic model assumes that the unmanned ship sails along the path at a preset speed and calculates the position coordinates at the future time by integration, while the dynamic model simulates the actual motion trajectory in combination with environmental parameters. The priority adjustment rule is a decision-making mechanism for determining the path adjustment object, which is implemented through the urgency of the task, the cost of the path adjustment or a dynamic allocation mechanism. For example, the urgency of the task includes that high-priority tasks such as rescue ships maintain the original path, and the path adjustment cost involves the difference in energy consumption due to deviation from the original path. The possibility of the predicted trajectories intersecting can be determined by a space-time intersection detection algorithm. For example, it includes checking whether the grid cells of the two ship trajectories overlap at the same or adjacent time, or whether the predicted position spacing is less than the safety distance threshold.

[0094] Path adjustment is the process of correcting conflicting paths through local or global replanning. For example, local replanning involves reconnecting the endpoints after bypassing the conflict point, while global replanning involves recalculating the entire path to completely avoid the conflict area. During the adjustment process, the cost model must be re-evaluated to evaluate the comprehensive cost of the new path to ensure a balance between safety and efficiency. The process of recalculating predicted trajectories and repeatedly checking them can be an iterative verification mechanism. For example, after adjusting a ship's path, its trajectory must be rechecked for conflicts with other ships in the formation until all ship pairs' trajectories have no possibility of intersecting.

[0095] It should be noted that the judgment of the predicted trajectories of the two unmanned ships specifically includes: obtaining the initial planned path, maximum navigation speed, minimum turning radius and current position information of any two unmanned ships in the unmanned ship formation, and calculating the predicted position coordinates of the two unmanned ships at several future time nodes based on a preset time interval; based on the predicted position coordinates of the two unmanned ships at several future time nodes, calculating the distance between the predicted positions of the two unmanned ships at the same time node. If the distance is less than the preset safety distance threshold, and the predicted trajectories of the two unmanned ships before and after the time node show a trend of gradually approaching, it is determined that there is a possibility that the predicted trajectories of the two unmanned ships may intersect; wherein, the safety distance threshold is dynamically adjusted according to the hull length, width and navigation speed of the unmanned ship.

[0096] The maximum navigation speed is the highest speed of the unmanned ship at rated power, which is usually determined by the performance of the power system and obtained through the ship parameter database or real-time sensor feedback. For example, it is a rated speed limit of 5m / s. The minimum turning radius is the minimum curvature radius of the unmanned ship when turning at the rated speed. It is determined by the ship's rudder angle limit and the power system response characteristics. It can be determined by the ship's kinematic parameters or experimental calibration data. For example, it is a minimum turning radius constraint of 8 meters. The current position information is the coordinate data output by the unmanned ship's real-time positioning system, which is collected by sensors such as satellite navigation or inertial measurement units. For example, it includes longitude and latitude coordinates or two-dimensional coordinate values ​​in a local coordinate system. The preset time interval can be the sampling period of the cloud platform or control system, which is set through the system configuration parameters. For example, it is a fixed time step of 10 seconds.

[0097] The predicted position coordinates are the result of mathematically deducing the theoretical position of the unmanned vessel at a future moment based on the current state and motion model. The calculation methods include kinematic models or dynamic models. The kinematic model assumes that the unmanned vessel sails along the path at a constant speed and calculates the displacement by integration. The dynamic model combines acceleration limits, water flow effects, and steering inertia to simulate a more realistic motion trajectory. For example, the curvature radius ≥ minimum turning radius constraint must be met in the turning section of the path. The safety distance threshold is the minimum safe distance allowed between the two ships. Its dynamic adjustment formula can be expressed as: safety distance = k × (hull length + hull width) + c × relative speed, where k and c are empirical coefficients. For example, when the relative speed of the two ships is 10m / s, the coefficient c may be taken as 0.5 to increase the safety distance.

[0098] This invention builds a prediction model by acquiring the initial planned path, navigation parameters, and real-time position information. It then uses segmented time nodes to calculate predicted positions and combines them with dynamic safety distance thresholds for distance and trend analysis, achieving the technical effect of improving the accuracy and adaptability of trajectory conflict detection. This solution uses physical constraints to prevent misjudgments of infeasible paths, utilizes a dynamic threshold mechanism to adapt to different ship types and speed scenarios, and filters out false alarms of brief approaches through trend analysis. Ultimately, this creates a closed-loop optimization with digital maps, cost models, and trajectory adjustment modules, significantly enhancing the safety of multi-vessel formation navigation and the robustness of path planning.

[0099] In one embodiment of the present invention, the adjustment of the initial optimal planned path of one of the unmanned ships according to the priority specifically includes: determining the unmanned ship whose optimal planned path needs to be adjusted based on the task priority, the current load status and the remaining distance from the target position; re-selecting a path from multiple candidate paths whose cost function value difference with the original path is within a preset range and can avoid the risk of collision; if there is no candidate path whose cost function value difference with the original path is within the preset range, locally optimizing the original optimal planned path, and adding a turning point near the predicted collision area during the local optimization; wherein, the setting of the turning point must meet the minimum turning radius constraint of the unmanned ship, and the path after turning is staggered in time and space with the predicted trajectory of the other unmanned ship.

[0100] The mission priority is a quantitative indicator that characterizes the importance of the mission currently being performed by the unmanned ship. It can be preset or dynamically adjusted through the mission management system. For example, the mission priority may include the priority of emergency rescue missions, environmental monitoring missions, and cargo transportation missions. The current load status is a set of parameters that reflect the real-time operating status of the unmanned ship. It can be obtained through sensor data acquisition or status monitoring modules. For example, the current load status includes the remaining battery power, cargo weight, and sensor working status. The remaining range from the target location can be the remaining distance from the current location of the unmanned ship to the mission endpoint, which can be calculated through the positioning system and digital map data. For example, the remaining range can be quantified in meters or nautical miles.

[0101] Candidate paths whose cost function values ​​differ from the original path within a preset range are considered to be a set of alternative paths that meet the path adjustment cost constraint. These paths are obtained through cost function calculation and threshold screening. Exemplarily, the preset range is set to ±10% to ±15% of the original path cost function value. Collision risk is the degree of spatial or temporal overlap between a path and the predicted trajectory of another unmanned vessel, evaluated using a trajectory prediction model. Exemplarily, collision risk can be determined based on a minimum safe distance threshold between the two vessel paths.

[0102] The process of determining which unmanned vessels require rerouting is accomplished through a cloud-based platform that comprehensively evaluates the multi-dimensional parameters of both conflicting parties. Specifically, the mission priorities are first compared, with higher-priority vessels maintaining their original paths. Secondly, the load status is analyzed. If a vessel's maneuverability is limited due to an overload, priority is given to adjusting another vessel. Finally, considering the remaining distance, vessels with a shorter remaining distance may retain their original paths to avoid mission delays. This process is implemented through weighted scoring or rule-based prioritization, ultimately determining the unmanned vessels requiring rerouting. This technical operation ensures that the path adjustment strategy strikes a balance between mission criticality, physical feasibility, and time constraints.

[0103] The process of selecting candidate paths can be achieved through cost difference constraints and collision avoidance effectiveness screening. Specifically, the cost function values ​​of multiple candidate paths are calculated to screen out paths that differ from the original path within a preset range. Secondly, the selected paths are verified to completely avoid the predicted collision area or achieve staggered traffic through spatiotemporal adjustments. If there are multiple paths that meet the requirements, the path with the lowest cost is selected as the final path. This technical operation ensures that path adjustments strike a balance between safety and efficiency, avoiding a decrease in overall mission performance due to excessive deviation from the original path.

[0104] The process of locally optimizing the path can be achieved through the insertion of turning points and the satisfaction of constraints. Specifically, this involves inserting turning points near the predicted collision zone and generating smooth path segments using trajectory planning algorithms such as cubic spline interpolation; calculating turning point coordinates to ensure that the path curvature radius is no less than the minimum turning radius; and adjusting the turning point position or speed to ensure that the time interval between the two ships passing through the collision zone meets the preset requirements. This technique provides local corrections when candidate paths are insufficient, ensuring physical feasibility while avoiding the high computational cost of re-planning the entire path.

[0105] The present invention achieves the following technical effects through multi-dimensional parameter evaluation to determine path adjustment targets, cost constraint screening of candidate paths, and local path optimization and constraint satisfaction: an intelligent decision-making mechanism prioritizes low-priority or high-load vessels, reducing interference with critical tasks; path adjustment costs are kept within a reasonable range to avoid significant increases in mission time; turning point constraints ensure the physical feasibility of the path and prevent power system overload; and a time-space staggered strategy enables orderly passage of formation vessels through conflicting areas. These technical approaches work synergistically to significantly improve the rationality, reliability, and dynamic adaptability of path adjustment, particularly in complex multi-vessel formation scenarios, effectively coordinating navigation conflicts while ensuring the efficiency of mission execution and navigation safety for each ship.

[0106] In one embodiment of the present invention, updating the digital map using environmental information specifically includes: updating the traffic status of corresponding grid cells in the digital map based on newly detected obstacle information, marking the grid cells where the newly appeared obstacles are located as impassable cells; updating the water flow impact parameters of each grid cell in the digital map based on new water flow speed and direction information; and re-using a heuristic search algorithm to generate multiple candidate paths, using the grid cell corresponding to the current position of the unmanned boat as a new starting point and the grid cell corresponding to the target position as an end point, to re-screen the optimal planned path.

[0107] Among them, the traffic status refers to the binary mark of whether the grid unit is passable, which is obtained by mapping the coordinates of the obstacle detected by the unmanned ship sensor and the coordinates of the digital map grid. Exemplarily, the traffic status includes the extended no-entry area mark based on the safety boundary. The water flow influencing parameters include the water flow velocity vector, which is obtained through the flow velocity and direction meter carried by the unmanned ship or the external hydrological data interface. Exemplarily, the water flow parameters include real-time updated water flow vector field data. After receiving the environmental data reported by the unmanned ship through the cloud platform, spatial mapping is first performed to convert the coordinates into grid indexes, and then the traffic status and water flow vector of the corresponding grid unit are updated. Finally, the heuristic search algorithm is re-executed with the current ship position as the starting point to generate a set of candidate paths, thereby achieving the technical effect of dynamically avoiding obstacles, adapting to water flow changes, and ensuring the real-time status of path connection.

[0108] The present invention also provides an unmanned ship formation path planning system, comprising a processor and a memory for storing instructions executable by the processor; wherein the processor is configured to execute the instructions to implement the steps of the cloud-controlled unmanned ship formation path planning method as described above.

[0109] The processor refers to a computing unit in a cloud platform or edge computing node, and its functions include executing path planning algorithms and cost model calculations. Multi-threaded task processing can be achieved through a parallel computing architecture. For example, the processor may include a multi-core CPU that supports the SIMD instruction set or a GPU configured with CUDA acceleration. The memory is used to persistently store digital map data, sensor information, and intermediate results of the algorithm. For example, the memory may include a hybrid storage system that combines a cache and an SSD. The processor loads the digital map and algorithm code in the memory, and multi-threaded parallel processing of trajectory prediction and conflict detection for different ships is used when executing multi-ship path planning tasks. At the same time, the hierarchical storage architecture is used to accelerate the read and write operations of high-frequency access data, thereby achieving the technical effect of improving the concurrent processing capability of large-scale formation path planning and reducing system latency.

[0110] The present invention updates the traffic status and water flow parameters of the digital map in real time to reflect environmental changes, combines it with heuristic path replanning based on the current ship position to ensure the real-time status of path connection, and relies on processor parallel computing and hierarchical storage architecture to achieve efficient execution of high-concurrency path planning tasks. It can significantly improve the timeliness and task reliability of the system's path planning in dynamic and complex environments, and support large-scale formations to maintain low-latency response and global optimal path generation in scenarios such as sudden obstacles and sudden water flow changes.

[0111] The above description is only a preferred embodiment of the present invention. It should be pointed out that ordinary technicians in this technical field can make several improvements and modifications without departing from the principles of the present invention. These improvements and modifications should also be regarded as the scope of protection of the present invention.

Claims

1. A cloud-controlled unmanned vessel formation path planning method, characterized in that: The following steps are involved: Obtain the initial position information, target position information and environmental information of each unmanned ship in the unmanned ship formation through the cloud platform; Based on environmental information, a digital map of the navigation area is constructed through the cloud platform; Based on the initial position information and target position information of each unmanned vessel, multiple candidate paths are planned for each unmanned vessel through a digital map; Based on multiple candidate paths, the cost function value of each candidate path is calculated through a preset cost model, and the initial optimal planning path is screened out based on the cost function values ​​of the multiple candidate paths; During the navigation process of each unmanned vessel according to the initial optimal planned path, the cloud platform receives its own position information and environmental information in the navigation area in real time, updates the digital map based on the environmental information, and updates the optimal planned path based on the updated digital map and its own position information.

2. The cloud-controlled unmanned vessel formation path planning method according to claim 1, characterized in that: The environmental information includes obstacle distribution information, water flow speed and direction information. The digital map divides the navigation area into multiple grid cells, each grid cell corresponds to a unique coordinate. The digital map of the navigation area constructed through the cloud platform specifically includes: Based on the obstacle distribution information, the grid cells containing obstacles are marked as impassable cells through the cloud platform; Based on the water flow speed and direction information, the corresponding water flow impact parameters are assigned to each grid unit through the cloud platform.

3. The cloud-controlled unmanned vessel formation path planning method according to claim 2, characterized in that: The method of planning multiple candidate paths for each unmanned vessel using a digital map specifically includes: A heuristic search algorithm is used, with the grid cell corresponding to the initial position of the unmanned vessel as the starting point and the grid cell corresponding to the target position as the end point, to search and generate multiple paths leading to the end point on the digital map as candidate paths; During the search process, ensure that the candidate path does not pass through the grid cells marked as impassable cells.

4. The cloud-controlled unmanned vessel formation path planning method according to claim 1, characterized in that: The cost function value of each candidate path calculated by the preset cost model specifically includes: Calculate the number of grid cells that the candidate path passes through and determine the path length cost; Calculate the distance between the center of each grid cell on the candidate path and the nearest obstacle, accumulate the inverse of the distance, and determine the distance cost to the obstacle; Based on the water flow impact parameters and the angle between the unmanned boat's navigation direction and the water flow direction, the water flow impact component in each grid cell is calculated. The water flow impact components in all grid cells on the candidate path are accumulated to determine the water flow impact cost. The path length cost, obstacle distance cost and water flow impact cost are weighted and summed according to the preset weight coefficient to obtain the cost function value of the candidate path.

5. The cloud-controlled unmanned vessel formation path planning method according to claim 2, characterized in that: The real-time receiving of the position information of each unmanned vessel and the environmental information within the navigation area through the cloud platform specifically includes: The unmanned boat obtains its own location information through its own positioning equipment and sends its location information to the cloud platform through the wireless communication module; The unmanned boat uses its own sensors to detect new obstacles in the navigation area and changes in water speed and direction, and sends the obstacle information and change information to the cloud platform through the wireless communication module.

6. The cloud-controlled unmanned vessel formation path planning method according to claim 1, characterized in that: The process of each unmanned vessel navigating according to the initial optimal planned path also includes the following steps: Based on the initial optimal planning path of each unmanned ship in the unmanned ship formation, the predicted trajectories of any two unmanned ships in the future are calculated and the predicted trajectories of the two unmanned ships are judged; If there is a possibility of intersection between the predicted trajectories, the initial optimal planned path of one of the unmanned ships is adjusted according to the priority. Based on the adjusted optimal planned path, the adjusted unmanned ship is recalculated to see whether there is a possibility of intersection between the predicted trajectories of other unmanned ships in the unmanned ship formation. If there is a possibility of intersection, the optimal planned path of the unmanned ship is readjusted until there is no possibility of intersection between the predicted trajectories of the adjusted unmanned ship and other unmanned ships in the unmanned ship formation. If not, other pairs of unmanned ships are continued to be checked until there is no possibility of intersection between the predicted trajectories of all unmanned ships.

7. The cloud-controlled unmanned vessel formation path planning method according to claim 6, characterized in that: The judgment of the predicted trajectories of the two unmanned ships specifically includes: Obtain the initial planned path, maximum navigation speed, minimum turning radius and current position information of any two unmanned ships in the unmanned ship formation, and calculate the predicted position coordinates of the two unmanned ships at several future time nodes based on the preset time interval; Based on the predicted position coordinates of the two unmanned ships at several future time points, the distance between the predicted positions of the two unmanned ships at the same time point is calculated. If the distance is less than the preset safety distance threshold and the predicted trajectories of the two unmanned ships before and after the time point show a trend of gradually approaching, it is determined that there is a possibility that the predicted trajectories of the two unmanned ships intersect; The safety distance threshold is dynamically adjusted according to the length, width and navigation speed of the unmanned ship.

8. The cloud-controlled unmanned vessel formation path planning method according to claim 7, characterized in that: The step of adjusting the initial optimal planning path of one of the unmanned ships according to the priority specifically includes: Determine the unmanned vessel that needs to adjust the optimal planned path based on mission priority, current load status, and remaining distance to the target location; Reselect a path from multiple candidate paths whose cost function value is within a preset range and can avoid collision risks; If there is no candidate path whose cost function value differs from the original path within the preset range, the original optimal planned path is locally optimized, and turning points are added near the predicted collision area during local optimization; Among them, the setting of the turning point must meet the minimum turning radius constraint of the unmanned ship, and the path after turning must be staggered in time and space with the predicted trajectory of another unmanned ship.

9. The cloud-controlled unmanned vessel formation path planning method according to claim 5, characterized in that: The updating of the digital map by using environmental information specifically includes: Based on the newly detected obstacle information, the traffic status of the corresponding grid cell in the digital map is updated, and the grid cell where the new obstacle is located is marked as an impassable cell; Update the water flow impact parameters of each grid cell in the digital map based on the new water flow speed and direction information; Taking the grid cell corresponding to the current position of the unmanned ship as the new starting point and the grid cell corresponding to the target position as the end point, the heuristic search algorithm is re-adopted to generate multiple candidate paths and the optimal planning path is re-screened.

10. The cloud-controlled unmanned vessel formation path planning system is characterized by: comprising a processor and a memory for storing instructions executable by said processor; The processor is configured to execute the instructions to implement the steps of the cloud-controlled unmanned ship formation path planning method as described in any one of claims 1 to 9.

Images