A communication window-aware AUV hierarchical path planning method and system

Abstract

This invention belongs to the field of offline mission and path planning technology for autonomous underwater vehicles (AUVs), and relates to a hierarchical path planning method and system for AUVs with communication window awareness. The method includes: generating a set of spatiotemporal communication anchor points for multimodal communication resources based on input information; constructing a spatiotemporal planning network to obtain a spatiotemporal expansion graph; constructing a multi-objective optimization function and searching for non-dominated solution sets on the spatiotemporal expansion graph using an improved label correction algorithm to obtain a macroscopic path skeleton; obtaining a refined executable trajectory based on the macroscopic path skeleton, ocean current field, and dynamic constraints using an improved ATW algorithm with spatiotemporal dual-constraint pruning; and outputting a complete planning product including a mission plan, cruise path, and communication plan. This invention enables heterogeneous communication opportunities to be jointly solved within a single planning framework. Through multi-objective hierarchical optimization and spatiotemporal dual-constraint pruning, it ensures that the arrival time at the communication anchor point strictly falls within the available time window, thus solving various defects in existing communication event processing solutions.

Description

Technical Field

[0001] This invention belongs to the field of offline mission and path planning technology for autonomous underwater vehicles, specifically relating to a communication window-aware AUV hierarchical path planning method and system. Background Technology

[0002] During long-range operations, Autonomous Underwater Vehicles (AUVs) are constrained by the complex underwater environment, making communication between AUVs and shore stations highly dependent on limited and heterogeneous link resources. Currently, commonly used communication link resources mainly fall into four categories, with differences in operating conditions and applicable scenarios for each type: 1) Iridium satellite short burst data (SBD) link: AUVs need to float to the surface and can only establish a communication connection within the visible window when the satellite passes overhead. The typical duration of a single satellite overhead is between 30 seconds and 10 minutes. Multiple satellites pass overhead alternately to form a complex discrete time window sequence. 2) UHF (or VHF) radio link: The AUV must float to the surface and the distance between the AUV and the mother ship or base station must not exceed the line-of-sight communication range (typical effective distance 5km to 30km). 3) Underwater acoustic relay link: AUV does not need to surface, but it is required to maintain a certain distance and depth relationship with the underwater relay buoy or the transducer deployed by the mother ship. Its effective communication radius is significantly affected by the acoustic field conditions. 4) Wi-Fi link (or 4G link): AUVs need to float to the surface and can only be used in special surface conditions near the shore or near the mother ship, with a very small coverage area.

[0003] The AUV shore-based command and control system is responsible for mission planning and path generation before AUV operations. Its planning outputs (mission plan, cruise path, communication plan) directly determine the mission coverage completeness, energy consumption level, and communication transmission integrity of the AUV during autonomous underwater operations. A review of existing literature reveals that current AUV path planning research mainly follows these technical routes: path planning based on deterministic search, path planning based on random sampling, path planning based on swarm intelligence, energy-optimal planning combined with ocean current fields, planning considering communication propagation loss, multi-AUV collaborative planning, and planning and positioning considering stealth.

[0004] The aforementioned technical approaches have relatively mature solutions in terms of mission integrity, energy consumption, and obstacle avoidance. However, the handling of communication events typically employs the following simplified methods: a) periodically inserting ascent and return points along the path at fixed times or distances; b) setting the ascent points as several manually designated fixed waypoints; c) calculating the ascent timing using an independent module after path planning is complete; d) incorporating communication link quality (such as acoustic communication propagation loss) into the objective function without distinguishing specific time windows. However, these simplified methods for handling communication events have the following drawbacks: 1) Simplified communication resource modeling leads to window mismatch: Traditional methods mostly insert floating points at fixed time periods (such as every 3 or 4 hours), which cannot guarantee that the satellite will be visible when the AUV reaches the water surface; Iridium satellite is a low-orbit non-geostationary satellite, and the overhead window of a single satellite at a specific geographical point is on the order of minutes. Multiple satellites alternately pass overhead to form a complex discrete time window sequence; window mismatch will lead to the situation where the AUV cannot establish a connection after rising and is forced to dive after wasting energy, which wastes power and delays communication backhaul; 2) Decoupling path and time solution: Mainstream planners generally generate spatial paths first, and then calculate the arrival time with constant or approximate speed; however, due to factors such as ocean current disturbances, obstacle avoidance and detour, and profile operations, there is a significant cumulative error in the arrival time of AUVs, which leads to a mismatch between the planned ascent window and the actual arrival time, with errors reaching as high as 10 to 30 minutes. 3) Lack of unified modeling for multiple communication methods: The availability conditions of the four types of links, namely Iridium, radio, underwater acoustic relay, and Wi-Fi, are very different (Iridium depends on time window, radio depends on distance, underwater acoustic relay depends on relative position, and Wi-Fi depends on near field). Existing planning methods usually only deal with one type of communication link, or only use the propagation loss of underwater acoustic communication as the only scalar indicator, which cannot achieve unified utilization of multimodal communication opportunities under a single planning framework. 4) Insufficient coupling of multiple objectives: The integrity of task coverage, energy consumption level and communication backhaul integrity are mutually constrained, but most existing planning methods take energy consumption level or path length as a single objective. A few multi-objective planning methods focus on path length + underwater acoustic propagation loss and lack special modeling for communication backhaul integrity. 5) Lack of emergency rollback mechanism: When the preset buoy point fails to match the satellite window, the existing system either retryes or continues to dive and waits for the next cycle. It does not have the ability to switch online to alternative communication anchor points (such as another satellite passing overhead, underwater relay buoy, or radio line of sight range), resulting in poor mission robustness. 6) High operator burden: The key parameters of existing planning methods, such as multi-objective weights and floating period, mostly need to be manually adjusted by operators for specific tasks, which depends on operator experience and results in poor consistency of planning results. Summary of the Invention

[0005] In view of the shortcomings of related technologies, the present invention provides a communication window-aware AUV hierarchical path planning method and system to solve at least one of the technical problems mentioned in the background art.

[0006] This invention provides a communication window-aware AUV hierarchical path planning method, comprising the following steps: S1. Obtain input information, including mission intent and constraints, satellite ephemeris, mother ship or base station navigation plan, ocean current and environmental forecasts, AUV parameters and operating conditions; S2. Generate a set of spatiotemporal communication anchor points based on the input information. , No. Communication anchor points Represented as equation (1); where, , , , , , and The first The planar coordinates of each communication anchor point, the depth range of the required AUV, the time window, the link type, the estimated available bandwidth, the reliability estimate, and the buoyancy risk score. The link types include satellite links, radio links, underwater acoustic relay links, and Wi-Fi links. (1); S3. Construct a spatiotemporal planning network, including: constructing a spatial base map of the task region according to grids or topological key points. ;right Each node in the dataset is time-extended along a discretized time axis corresponding to the AUV's expected arrival time, resulting in a spatiotemporal node set. For each Generate communication event nodes and will and Connect all nodes that satisfy the triple matching of space, time, and depth to obtain a spatiotemporal extended graph. ; S4. Perform global hierarchical planning, including: Constructing a multi-objective optimization function , expressed as equation (3); where, for An effective path on the way, , and These are the costs associated with task coverage integrity, energy consumption, and communication backhaul integrity, respectively. , and These are, respectively, the task coverage integrity weight, energy consumption weight, and communication return integrity weight driven by the working condition template; (3); To minimize With the goal of, in An improved label correction algorithm is used to search the Pareto non-dominated solution set to obtain the macroscopic path skeleton. , This includes the primary communication anchor sequence and coarse-grained arrival time windows; S5. Perform local dynamic refinement, including: For the target skeleton, an improved A algorithm with spatiotemporal dual-constraint pruning is adopted. The TW algorithm solves the sub-path segment by segment, introduces an ocean current compensation model to correct the actual arrival time of the AUV, and obtains a fine-grained executable trajectory through dual verification of spatial reachability and temporal reachability. ; S6. Output complete planning deliverables, including mission plan, cruise route, and communication plan.

[0007] In some embodiments, an emergency handover strategy generation step is also included, which includes: based on And the sequence of primary communication anchor points, setting the latest decision time for each primary communication anchor point. Within the preset time neighborhood of each primary communication anchor point, multiple alternative communication anchor points of the same and different types are searched and sorted to obtain a list of alternative communication anchor points, thereby constructing a primary + alternative communication anchor point decision tree. When the AUV is operating online, the primary communication anchor point is selected first. If it is determined before the latest decision time that the AUV cannot reach the primary communication anchor point, the alternative communication anchor points in the alternative communication anchor point list are tried in sequence until the first feasible alternative communication anchor point is hit.

[0008] In some embodiments, the latest decision time for each primary communication anchor point in the emergency handover strategy generation step. According to equation (9), where, For the first The start time of each communication anchor point time window For decision-making and switching margin, it is calculated according to equation (10). The shortest time required to determine AUV deviation. The time required for switching to an alternative communication anchor point, For safety margin; (9); (10); The sorting strategies for multiple candidate communication anchor points include ascending order based on the travel cost of switching from the primary communication anchor point to a candidate communication anchor point, descending order based on the reliability estimate of the candidate communication anchor points, and ascending order based on the time proximity of the candidate communication anchor points. The top-ranked candidates are then selected. An ordered linked list of alternative communication anchor points is constructed from the available alternative communication anchor points. .

[0009] In some embodiments, in step S3, the spatial base map , for The set of nodes, for The set of edges; for Each node in Discretized timeline based on AUV's estimated arrival time Perform time extension to obtain a set of spatiotemporal nodes. , expressed as equation (2); where, For a moment, At the starting time, The time spent walking is long; (2); Spatial, temporal, and depth triple matching includes: nodes With communication anchor point The distance between them does not exceed the preset distance threshold and time. Landing at the communication anchor point Time window Indoor and AUV can be used at any time Reaching depth range ; Spacetime Extension Graph , for The set of vertices, , for The set of edges, ;in, For the reason Derived spacetime edges, for Nodes that satisfy spatial, temporal, and depth matching The edges connecting them.

[0010] In some embodiments, in step S4, the task coverage completeness cost is... Represented as equation (4), Assigning weights to each task. For the task Completion rate; energy consumption cost Represented as equation (5), AUV speed relative to water The following sailing power, for The edges in For AUV along The sailing time Additional energy consumption during AUV ascent or descent. and These are the lower and upper limits of energy consumption normalization; and the cost of communication backhaul integrity. Represented as equation (6), Preset priority weights for communication feedback during the task phase; (4); (5); (6).

[0011] In some embodiments, step S4 further includes optimizing the multi-objective function. Set energy constraints, buoyancy feasibility constraints, and weight constraints; the energy constraint is as follows: ,in, For AUV along Energy consumption during navigation Available power for AUV For battery margin, The feasibility constraint for surfacing is that the AUV, at any spatiotemporal node of any surfacing segment... Capable of rising to The dynamic feasibility; the weight constraint condition is , , and All are greater than 0.

[0012] In some embodiments, in step S4, the operational mode template includes long-range survey, emergency search, energy-saving cruise, communication support, and multi-mission integration; when the AUV is in the long-range survey operational mode, , , When the AUV is in emergency search and rescue operation, , , 0.6; When the AUV is in energy-saving cruise mode, , , When the AUV is in a communication support mode, , , When the AUV is in a multi-tasking operating condition, , , .

[0013] In some embodiments, the correction of the actual arrival time of the AUV in step S5 includes: Establish the current compensation model shown in Equation (7) to calculate the actual ground velocity of the AUV. ;in, Here are the position coordinates of the AUV. For a moment, The water velocity of the AUV, For ocean current vector field; (7); For each edge The actual arrival time of the AUV is solved iteratively according to equation (8). ,in, for Location of the AUV at any given time For heading; (8).

[0014] In some embodiments, in step S5, the improved A with spatiotemporal dual-constraint pruning... The TW algorithm includes the following sub-steps: S51, Read in the sub-path starting point, target communication anchor point and its time window, and ocean current field; S52. Add the starting point of the sub-path to the OPEN list, set its cumulative cost to 0, and set its estimate to heuristic estimate. S53. Retrieve the OPEN state with the lowest estimated value from the OPEN list. ,in, For planar position, For the course, The time when this state is reached; determine Has the target communication anchor point been reached? If yes, backtrack the parent pointer, output the fine sub-path, and the algorithm terminates; otherwise, proceed with... Perform neighborhood expansion and retrieve the state of an unprocessed neighbor. ; S54, to Perform spatial reachability verification to determine Do they satisfy simultaneously? Located within the mission domain and by arrive Satisfy AUV dynamic constraints; otherwise, then Discard and return to sub-step S53 to continue execution; if so, continue with... Perform time window reachability verification to determine Can it arrive within the time window of the target communication anchor point? If not, then... Discard and return to sub-step S53 to continue execution; if so, calculate. Actual arrival time and transit time arrive The new cumulative cost; S55. Determine if the new cumulative cost is less than... If the recorded cost is not found, return to step S53 to continue execution; if so, update... The cumulative cost, valuation, and parent pointer, and will Add to the OPEN list and return to step S53 to continue execution; the algorithm terminates when the fine sub-paths are output or the OPEN list is exhausted.

[0015] This invention also provides a communication window-aware AUV hierarchical path planning system for executing the aforementioned communication window-aware AUV hierarchical path planning method; the communication window-aware AUV hierarchical path planning system includes: The input module is used to obtain input information including mission intent and constraints, satellite ephemeris, mother ship or base station navigation plan, ocean current and environmental forecasts, AUV parameters and operating conditions; The communication resource modeling module is used to generate a set of spatiotemporal communication anchor points based on input information; The spatiotemporal planning network construction module is used to obtain a spatiotemporal extension graph based on the set of spatiotemporal communication anchor points and the task space description; The global hierarchical planning module is used to construct a multi-objective optimization function and use an improved label correction algorithm to search for Pareto non-dominated solution sets on the spatiotemporal extension graph to obtain the macroscopic path skeleton. The local refinement module is used to improve the AUV's performance by employing spatiotemporal dual-constraint pruning based on the macroscopic path framework, ocean current field, and AUV dynamic constraints. The TW algorithm obtains a fine-grained executable trajectory; The emergency anchor point rollback module is used to obtain the latest decision time of each primary communication anchor point and the alternative communication anchor point list, and construct a primary + alternative communication anchor point decision tree; The output module is used to output complete planning deliverables, including mission plans, cruise routes, communication plans, and emergency handover strategies.

[0016] Based on the above technical solutions, the communication window-aware AUV hierarchical path planning method and system in this embodiment of the invention achieves unified modeling of various heterogeneous communication resources through STCA seven-tuples. During the planning phase, the available multimodal communication window is incorporated as a spatiotemporal constraint into AUV path optimization, enabling heterogeneous communication opportunities to be jointly solved within a single planning framework. Relying on the spatiotemporal extension graph, communication opportunities are treated as optional branches of path nodes in the time dimension, achieving spatiotemporal binding between communication events and path nodes, overcoming problems such as communication window mismatch and path-time decoupling in traditional solutions; a global approach is adopted. The locally layered three-objective optimization architecture balances task coverage completeness, energy consumption, and communication feedback completeness. Combined with spatiotemporal dual-constraint pruning, it improves search and solution efficiency and ensures trajectory dynamics feasibility. It utilizes operational condition templates to achieve adaptive configuration of multi-objective weights, eliminating reliance on operator experience for manual parameter tuning and improving the consistency of planning results. Based on the latest decision time and the primary + alternative communication anchor point decision tree, it pre-generates emergency switching strategies to cope with various disturbances during execution, enhancing task robustness. This invention fully utilizes multimodal communication opportunities, reducing energy loss and feedback delays caused by ineffective surfacing, improving data feedback quality and task completion levels during AUV autonomous underwater navigation operations. It can be applied to long-range marine surveys, underwater target search, seabed topography mapping, underwater environmental monitoring, and can also be used as a single-unit plan in future multi-AUV collaborative planning. Attached Figure Description

[0017] The accompanying drawings, which are included to provide a further understanding of the invention and form part of this application, illustrate exemplary embodiments of the invention and, together with their description, serve to explain the invention and do not constitute an undue limitation thereof. In the drawings: Figure 1 A flowchart of the communication window-aware AUV hierarchical path planning method of the present invention; Figure 2 This is a diagram showing the overall architecture of the communication window-aware AUV hierarchical path planning system of the present invention. Figure 3 This is a data flow diagram of the global hierarchical planning module and the local refinement module in this invention; Figure 4 An improvement A to the spatiotemporal dual-constraint pruning method in this invention. Flowchart of the TW algorithm; Figure 5 This is an example diagram of the primary + alternative communication anchor point decision tree in this invention. Detailed Implementation

[0018] The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of them. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are within the scope of protection of the present invention.

[0019] In the description of this invention, it should be understood that the terms "center", "lateral", "longitudinal", "upper", "lower", "top", "bottom", "inner", "outer", "left", "right", "front", "rear", "vertical", "horizontal", etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are only for the convenience of describing this invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on this invention.

[0020] In the description of this invention, it should be noted that, unless otherwise explicitly specified and limited, the terms "installation," "connection," and "joining" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a direct connection or an indirect connection through an intermediate medium; and they can refer to the internal communication between two components. Those skilled in the art can understand the specific meaning of the above terms in this invention based on the specific circumstances.

[0021] refer to Figures 1-4 As shown, the present invention provides a communication window-aware AUV hierarchical path planning method, including the following steps S1 to S6.

[0022] Step S1: Obtain input information, including mission intent and constraints, satellite ephemeris (TLE), mother ship or base station navigation plan, ocean current and environmental forecasts, AUV parameters and operating conditions.

[0023] Step S2: Generate a set of spatiotemporal communication anchor points based on the input information. , No. Communication anchor points Defined in the form of a seven-tuple, it is expressed as equation (1); (1); In equation (1), For the first The planar coordinates of the communication anchor points , used to indicate the geographical location of the communication anchor point; For the first The depth range of the AUV required for each communication anchor point represents the effective communication / sensing depth range of the communication anchor point, constraining spatial coverage capability. , and The first Minimum and maximum depth of AUV required for each communication anchor point; For the first The time window of each communication anchor point represents the effective working interval of the communication anchor point in the time dimension. and The first The start and end times of each communication anchor point; For the first The link type for each communication anchor point includes satellite links (abbreviated as...) ), radio link (abbreviation) ), underwater acoustic relay link (abbreviated as) Wi-Fi link (abbreviated) ), For discrete indices, This refers to communication links that cover four types of communication anchor points; For the first Estimated available bandwidth for each communication anchor point; For the first Reliability estimation (i.e., probability of successful transmission) for each communication anchor point. ; For the first Risk score for the upward movement of each communication anchor point The typical values ​​of the seven-tuple for the four types of communication anchors are shown in Table 1.

[0024] Table 1 Typical values ​​of the seven-tuple for four types of communication anchor points

[0025] A brief theoretical explanation of the generation of various communication anchor points is provided. Iridium satellite anchor points: Based on the Iridium satellite two-line root (TLE) file, the center point of the mission area is determined within the planned time period. Perform satellite visibility forecasting; for each satellite, use the SGP4 orbit forecasting model to calculate its relative orbital visibility during the planned time period. Elevation sequence ;when (typical value) Visible periods are considered to be at an elevation angle of 8° or higher, meaning satellite communication can only be established when the elevation angle is large enough; visible periods are aggregated into one or more time windows. A satellite anchor point is generated for each time window. Radio anchor point: the expected location of the mothership or base station within the planned time period. Centered on, the range of sight Radio anchor points are generated with a radius of 10km to 30km (typical value); if the mother ship is a mobile vessel, sampling steps are taken according to its navigation plan. (Typical value 5 minutes) Discrete sampling, each sampling time Generate a time-limited anchor point. Underwater acoustic relay anchor point: based on the location of the underwater relay buoy or the suspended transducer. The effective radius of action is calculated based on the acoustic channel model, with the center as the reference point. (Typical values ​​3km to 10km) and corresponding effective depth zones (This refers to the depth band suitable for underwater acoustic communication, a common term in the field of acoustics); underwater acoustic relay anchors are characterized by more relaxed depth and location constraints, longer communication time windows but smaller bandwidth (typically hundreds to thousands of bits / s). Wi-Fi anchors: centered on the docking point or near the mother ship area, they have the highest bandwidth but extremely small coverage radius (typically 100m), and are usually used for high-bandwidth data backhaul at the end of a mission.

[0026] Step S3: Construct a spatiotemporal planning network; this step is used to invert communication opportunities into spatiotemporally coupled planning constraints, with the input being the task space description and the set of spatiotemporal communication anchor points. The output is a spatiotemporal extension graph. This step includes spatial base map construction (sub-step S31), time extension (sub-step S32), and communication event node binding (sub-step S33).

[0027] Sub-step S31: Construct a spatial base map of the task region according to grid or topological key points. , ,in, for The set of nodes, for The set of edges. Specifically, for the survey line task, This includes the two endpoints of all survey lines and possible turning points; for cruise missions, Including the origin, destination, and pre-defined waypoints; for search missions, It is composed of the center point of the grid. Each edge in Indicates from node To the node A navigable section of the waterway, in which, The two endpoints of this edge (waypoint / grid center / turn point); each edge is associated with 3 attributes: geometric distance. Estimated sailing time and estimated energy consumption ,in, Used to determine the time span of a spatiotemporal edge in a time extension. Used for energy consumption targets in subsequent step S4 Summation calculation.

[0028] Sub-step S32: For the spatial base map Middle node set Each node in the discretized time axis according to the AUV's expected arrival time Perform time extension to obtain a set of spatiotemporal nodes. , expressed as equation (2); where, For a moment, At the starting time, The time spent walking is long; (2).

[0029] The purpose of time extension is to give each spatial node a time dimension, allowing time to be explicitly considered as a state component in the planning search. (Time distance from walk length) The typical value ranges from 1 minute to 5 minutes, and can be adaptively adjusted according to the minimum width of the communication window; to ensure that no communication time window is "skipped", It must not exceed approximately half the width of the narrowest time window among all participating links. The values ​​for each link are based on Table 2; the system ultimately takes the minimum value required by each communication link, i.e. .

[0030] Table 2. Time distance from each link to the walk Reference table of values

[0031] Sub-step S33: Generate a set of spatiotemporal communication anchor points. Each of them Generate communication event nodes and will and All nodes that satisfy spatial, temporal, and depth triple matching Connect them to obtain a spacetime extension graph. .

[0032] Specifically, spatial, temporal, and depth triple matching includes: spatial matching is represented as... , i.e., node With communication anchor point The distance between them does not exceed the preset distance threshold. Time matching is represented as That is, time Landing at the communication anchor point Time window Internal; depth matching, AUV can be used at any time Reaching depth range .

[0033] Communication event node The weights are determined by The communication gain is determined, typically in the form of: ;in, This represents the priority weight for communication feedback during the current task phase.

[0034] After the above construction, a spatiotemporal extension graph is obtained. ; for The set of vertices, ; for The set of edges, ;in, For the reason The derived spacetime edge, specifically, for each space edge Generate from spatiotemporal nodes point to The directed edge, i.e., "at time..." from Departure and sailing time At arrive ”; for Nodes that satisfy spatial, temporal, and depth matching The edges connecting them.

[0035] Step S4: Perform global hierarchical planning, which includes the following sub-steps S41 to S42.

[0036] Sub-step S41: Construct a multi-objective optimization function , expressed as equation (3); where, For the spatiotemporal extension graph An effective path on the way, The starting point is the starting point of the task. The endpoint ∈ the target node set (such as the set of spatiotemporal nodes corresponding to the task endpoint or docking point). , and These are the costs of task coverage integrity, energy consumption, and communication backhaul integrity, respectively. , and These are, respectively, the task coverage integrity weight, energy consumption weight, and communication return integrity weight driven by the working condition template; (3).

[0037] Specifically, the cost of task coverage completeness Represented as equation (4); where, Assigning weights to each task. For the task Completion rate The survey line task is defined as the proportion of the survey lines that have been scanned out of the total survey lines. For sampling tasks, it's the percentage of visited sampling points; for search tasks, it's the area coverage of the search region. The more complete the task coverage, the higher the percentage of visited sampling points. The smaller the value, the more complete the task coverage. ; (4); Energy consumption cost Represented as equation (5), normalized to ;in, AUV speed relative to water The following sailing power, for The edges in For AUV along The sailing time Additional energy consumption during AUV ascent or descent. and These are the lower and upper limits of energy consumption normalization, respectively, and the minimum and maximum total energy consumption in the candidate path set are taken respectively (in engineering, the energy consumption of the shortest straight path can also be taken as ). Task power budget for The lower the energy consumption, The smaller the value; (5); Communication backhaul integrity cost Represented as equation (6); where, This sets a preset priority weight for communication feedback during the task phase; the more complete the communication feedback, the higher the priority weight. The smaller the value; This measure is used to evaluate the proportion of communication events that the planned path can accomplish relative to the task's expectations. It is considered successful when all expected communication events are reliably hit. ; (6).

[0038] Furthermore, for multi-objective optimization functions Set energy constraints, buoyancy feasibility constraints, and weight constraints. The energy constraint is as follows: ,in, For AUV along Energy consumption during navigation Available power for AUV For battery margin, , The typical value is 0.8, meaning a 20% reserve of power is provided to handle unforeseen circumstances and prevent the battery from being depleted during planning. The feasibility constraint for buoyancy is that, for any spatiotemporal node of the buoyancy segment... ,Require That is, the spatiotemporal node where the AUV is located in any upward segment. Capable of rising to The dynamic feasibility is determined (constrained by remaining battery power, current depth, and available ascent rate). The weighted constraints are as follows: , , and All are greater than 0.

[0039] Furthermore, the operational mode templates include long-range survey, emergency search, energy-saving cruise, communication support, and multi-mission integration. As shown in Table 3, when the AUV is in the long-range survey operational mode, , , When the AUV is in emergency search and rescue operation, , , 0.6; When the AUV is in energy-saving cruise mode, , , When the AUV is in a communication support mode, , , When the AUV is in a multi-tasking operating condition, , , The system automatically sets up according to the selected working condition template. , and No manual adjustment is required.

[0040] Table 3. Weight Adaptive Value Table Driven by Working Condition Template

[0041] Sub-step S42: Minimize the multi-objective optimization function With the goal of extending the spacetime graph An improved label correction algorithm is used to search the Pareto non-dominated solution set to obtain the macroscopic path skeleton. ; This includes the primary communication anchor sequence and coarse-grained arrival time window; the improved label correction algorithm itself is known to those skilled in the art and will not be elaborated here, but the innovation of this sub-step lies in: the state label not only contains the cumulative scalar cost, but the label also maintains a three-dimensional sub-cost vector ( , and This is to preserve the Pareto front during the search and avoid prematurely discarding solutions that are temporarily poor but globally non-dominated on a single objective.

[0042] Further explanation of the macroscopic path framework A clear definition: yes An ordered sequence of key spatiotemporal nodes, where each element is a tuple (mission waypoint or communication event node). (coarse-grained arrival time window); It specifies which communication anchors to access (i.e., which ones to hit). This includes the order in which communication anchors are accessed, the coarse-grained time / time window at which they arrive at each communication anchor, and the approximate path between communication anchors at the topology level; it is understandable that... It does not contain a refined, continuous trajectory that satisfies dynamics and current compensation; this part is generated by subsequent local dynamics refinement steps within each sub-path.

[0043] Step S5: Perform local dynamic refinement; which is based on For the target skeleton, an improved A algorithm with spatiotemporal dual-constraint pruning is adopted. The TW algorithm solves the problem segment by segment, refining the dynamic model, current compensation, and obstacle avoidance constraints within each segment of the path between adjacent skeletons / communication anchors. It introduces a current compensation model to correct the actual arrival time of the AUV and obtains a refined executable trajectory through dual verification of spatial and temporal reachability. Improvement A TW state is a quadruple These represent the position, heading, and time, respectively.

[0044] To further explain, the correction for the actual arrival time of the AUV includes: To accurately calculate the actual arrival time of the AUV, the ocean current compensation model shown in equation (7) is established to calculate the actual ground velocity of the AUV. ;in, Here are the position coordinates of the AUV. For a moment, The water velocity of the AUV, For ocean current vector field; (7); For each edge The actual arrival time of the AUV is solved iteratively according to equation (8). ,in, for Location of the AUV at any given time The heading is used; in actual implementation, discrete step numerical solution is used. (8).

[0045] To further explain, the improved A with spatiotemporal dual-constraint pruning The innovative idea behind the TW algorithm lies in simultaneously verifying both spatiotemporal reachability and time window reachability for each state to be expanded. States that fail to meet any verification condition are directly pruned, significantly reducing the search scale and saving computational power. The core data structure of this algorithm, the OPEN list, is A. The frontier state priority queue for TW-type search is prioritized according to the evaluation function. Sort by size from smallest to largest. ;in Starting point of the sub-path The cumulative actual cost to the current state (including navigation energy consumption / time). The current state is the communication anchor point to the target. The heuristic cost lower bound estimate can be obtained by using a linear reachability estimate that ignores ocean currents and obstacles. To reach this state Total cost estimation; the OPEN list stores "generated but not yet expanded" frontier states, which is the set of states to be expanded, not the target set; the algorithm always prioritizes expanding. The one with the smallest value. Specifically, refer to... Figure 4 As shown, the algorithm includes the following sub-steps S51 to S55.

[0046] Sub-step S51: Read in the starting point of the sub-path Target communication anchor point and its time window Ocean current vector field .

[0047] Sub-step S52: Set the starting point of the sub-path Add it to the OPEN list and set its cumulative cost to 0 (i.e. The valuation is set as a heuristic estimate (i.e.) The costs of the remaining states are considered as follows: The output of this step is an initial list of OPEN entries containing only the starting point.

[0048] Sub-step S53: Retrieve (or pop up) the valuation from the OPEN list. Minimum state to be expanded ,in, For planar position, For the course, The time when this state is reached. This is the expanded state in this round. (Judgment) Has the target communication anchor point been reached? (That is, satisfying the triple matching of space, time and depth); If so, then backtrack along the parent pointer, that is, traverse from the parent pointer... Back to the beginning The state sequence is obtained and reversed, thereby outputting a fine sub-path that satisfies the dynamic and time window constraints. The algorithm for this sub-path segment has successfully ended and terminated. If not, then... Perform neighborhood expansion by retrieving an unprocessed neighbor state according to the maneuver model. Then, proceed to the subsequent dual verification of spatiotemporal accessibility and time window accessibility.

[0049] Sub-step S54, for Perform spatial reachability verification to determine Do they satisfy simultaneously? Located within the mission domain (without crossing boundaries or entering obstacles / restricted areas) and by arrive Satisfy AUV dynamic constraints (turning radius, climb rate, speed limit, etc.); If not, then Deemed infeasible and discarded, i.e., for Prune the branches. It will no longer be added to the OPEN list, will not participate in subsequent expansions, and will return to sub-step S53 to continue execution, that is, to continue to retrieve the next state or process the next neighbor; If so, then continue with... Perform time window reachability verification; estimate by Continue to The earliest reachable time (using the lower bound of the most favorable ocean current / shortest time) and the latest reachable time (using the upper bound of the most unfavorable ocean current / longest time) are used to determine... Is it possible to achieve this within the time window of the target communication anchor point? Arrive within; if not (e.g., the earliest arrival is later than) Or the latest arrival is earlier than And if it cannot linger and wait), then the branch is unlikely to hit the communication anchor on time. If deemed infeasible, the process is discarded and returns to sub-step S53 to continue execution, i.e., to retrieve the next state or process the next neighbor; if so, the calculation is performed by referring to the ocean current compensation model. actual arrival time ( (Sailing time under ocean current compensation) and via arrive New cumulative cost The output is the candidate arrival time and candidate cost of the neighbor.

[0050] Sub-step S55: Determine the new cumulative cost Is it less than Recorded costs If not, abandon this update and return to sub-step S53 to continue execution; if yes, it means that the update has been completed. arrive Better, newer The cumulative cost ( ), valuation ( ), point the parent pointer to To update the parent pointer, and Add to the OPEN list (if it's already in the list, click New). (Adjust its position), and then return to sub-step S53 to continue execution, that is, continue to retrieve the next state or process the next neighbor.

[0051] It should be noted that before executing sub-step S53, it is necessary to check whether the OPEN list is empty; if it is empty, it means that there are no more scalable frontier states and the goal is unreachable, so the algorithm outputs "no feasible sub-path" and terminates; if it is not empty, then sub-step S53 is executed. Sub-steps S53 to S55 constitute the main loop; the algorithm for each sub-path segment terminates in two cases: one is hitting a communication anchor point, in which case a fine-grained sub-path is output. The algorithm terminates successfully if one path succeeds; otherwise, it terminates if the OPEN list is exhausted and no feasible sub-paths are output. If the algorithm terminates due to failure, a new global hierarchical planning process is required to update the macroscopic path skeleton. Then, perform local dynamic refinement again. All refined subpaths. After solving, all sub-paths are concatenated to output a fine-grained executable trajectory that satisfies the constraints of dynamics, obstacle avoidance, and communication time windows. .

[0052] Step S6: Output complete planning products, including mission plan, cruise route, communication plan, etc.

[0053] The above illustrative embodiment, by modeling multimodal communication resources in a unified spatiotemporal communication anchor point, enables heterogeneous communication opportunities such as Iridium satellite overhead, underwater acoustic relay, radio line-of-sight, and Wi-Fi near-field to be jointly solved under a single planning framework. Through the construction of a spatiotemporal extended planning graph, communication opportunities are inverted into strong spatiotemporal coupling constraints in the path planning stage, achieving spatiotemporal binding between communication events and path nodes. Through multi-objective hierarchical optimization and spatiotemporal dual-constraint pruning, macroscopic skeleton search and local dynamic refinement are decoupled, ensuring that the arrival time of the communication anchor point strictly falls within the available time window, while reducing the search scale and saving computing power. Through adaptive setting of multi-objective weights driven by operational condition templates, the operator's burden is significantly reduced, and planning consistency is improved. This embodiment can fully utilize multimodal communication opportunities, reduce energy loss and backhaul delays caused by ineffective surfacing, and improve the data backhaul quality and mission completion level during AUV autonomous underwater navigation operations.

[0054] To address deviations during the AUV execution phase after planning (such as actual ocean currents deviating from forecasts, mission delays, and reduced actual reachability of links), this invention pre-generates emergency switching strategies during the planning stage. (Reference) Figure 1 , Figure 5 As shown, in some embodiments, the communication window-aware AUV hierarchical path planning method further includes an emergency handover strategy generation step. This step includes: based on the macroscopic path skeleton... And the sequence of primary communication anchor points, setting the latest decision time for each primary communication anchor point. Within a preset time neighborhood (e.g., 30 minutes before and after) of each primary communication anchor point, multiple alternative communication anchor points of the same and different types are searched and sorted to obtain a linked list of alternative communication anchor points, thereby constructing a primary + alternative communication anchor point decision tree. The complete planning deliverable includes this emergency switching strategy. When the AUV is operating online, the primary communication anchor point is executed first. If it is determined that the primary communication anchor point is reachable before the latest decision time, the AUV proceeds to the primary communication anchor point as planned. If it is determined that the AUV cannot reach the primary communication anchor point before the latest decision time, it switches to alternative communication anchor points in the alternative communication anchor point linked list and tries them sequentially until the first feasible alternative communication anchor point is hit, at which point the AUV proceeds.

[0055] To further clarify, the latest decision time for each primary communication anchor point According to equation (9), where, For the first The start time of each communication anchor point time window For decision-making and switching margin, Calculated according to formula (10), The shortest time required for AUV to determine deviation (typically 30 seconds). The time required for switching to an alternative communication anchor point, For safety margin; The typical value is 3 min to 10 min. The meaning is: AUVs must be installed no later than [date missing]. Determine the primary communication anchor point before the time limit. Is it still reachable? If it is determined to be unreachable, then for the remaining time thereafter (i.e. arrive During this period, there is still time to flexibly and safely switch to an alternative communication anchor point; (9); (10).

[0056] The ranking strategy for multiple alternative communication anchor points includes sorting them in ascending order by the navigation cost (i.e., reachability cost) of switching from the primary communication anchor point to an alternative communication anchor point, and by reliability estimation of the alternative communication anchor points. Sort the candidate communication anchors in descending order, then in ascending order based on their time proximity, and take the top results of all sorted lists. An ordered linked list of alternative communication anchor points is constructed from the available alternative communication anchor points. It is understandable that It is a priority emergency chain, not a set that needs to be combined. The primary anchor point has the highest priority. For the same communication opportunity, only one communication anchor point is ultimately used, and its online selection semantics are sequential fallback; specifically, the AUV preferentially executes the primary communication anchor point. If in If the primary communication anchor point was previously determined to be unreachable, then switch to... The currently feasible and highest priority alternative communication anchor point ,like If that doesn't work, then try them one by one. , This continues until the first viable alternative communication anchor is hit or the linked list is exhausted; once a communication anchor is successfully executed, the current communication opportunity ends, and the remaining alternative communication anchors in the linked list are no longer used; therefore, it can be seen that the previous The candidate communication anchors are not combined in parallel, nor are they chosen arbitrarily. Instead, they are selected sequentially, with each communication opportunity ultimately landing on a single communication anchor (either the primary communication anchor or a candidate communication anchor). , It should be noted that while multiple communication anchors across different links may be utilized, the strategy for a single communication opportunity is always "one primary choice + one ordered alternative list, and finally, one communication anchor is executed".

[0057] refer to Figure 5 As shown, the emergency switching strategy for communication anchor points can be further refined: before the latest decision time, if it is determined that the deviation during AUV online operation does not exceed the preset value... At that time, the AUV proceeds to the primary communication anchor point as planned; if the deviation during AUV online operation is determined to be greater than the preset value... And did not exceed the preset value. When the AUV is in operation, it will travel to a neighboring communication anchor point of the same type as the primary communication anchor point (e.g., if the primary communication anchor point is a satellite anchor point, it will travel to another satellite anchor point); if the deviation of the AUV during online operation is determined to be greater than the preset value... When the AUV is in a different neighborhood than the primary communication anchor point (e.g., if the primary communication anchor point is a satellite anchor point, then it will go to a radio anchor point or an underwater acoustic relay anchor point); if the list is exhausted and no feasible alternative communication anchor point is found, an emergency safety alarm will be automatically triggered, the AUV will rise and the beacon will be activated.

[0058] The above illustrative embodiments, by setting the latest decision time of the primary communication anchor point and constructing a decision tree of primary and alternative communication anchor points, can output the complete emergency switching strategy required for the online execution phase of AUV during the planning stage, provide emergency rollback capability during the online execution phase, and have the ability to switch to alternative communication anchor points when the primary communication anchor point is unreachable, thereby improving the robustness of task execution.

[0059] refer to Figures 1-5 As shown, the present invention also provides a communication window-aware AUV hierarchical path planning system for executing the aforementioned communication window-aware AUV hierarchical path planning method. The communication window-aware AUV hierarchical path planning system includes an input module, a communication resource modeling module, a spatiotemporal planning network construction module, a global hierarchical planning module, a local refinement module, an emergency anchor point rollback module, and an output module; constituting a complete data flow from task intent input to complete planning product output.

[0060] The input module is used to obtain input information including mission intent and constraints, satellite ephemeris, mother ship or base station navigation plans, ocean current and environmental forecasts, and AUV parameters and operating conditions. The communication resource modeling module is used to generate a set of spatiotemporal communication anchor points based on the input information. The spatiotemporal planning network construction module is used to construct networks based on the spatiotemporal communication anchor point set. Obtain the spatiotemporal extension graph from the task space description Task space description refers to constructing a base map of the task space. The required mission geometry and topology information includes mission type identifier, mission geometric elements, mission area boundaries, start point, end point, obstacles and no-fly zones, depth / profile operation requirements, etc. The global hierarchical planning module is used to construct multi-objective optimization functions and extend the spatiotemporal graph. An improved label correction algorithm is used to search the Pareto non-dominated solution set to obtain the macroscopic path skeleton. This outputs the primary communication anchor point sequence and coarse-grained arrival time windows. The local refinement module is used to refine the macroscopic path skeleton. Constraints on ocean current field and AUV dynamics, and an improved AUV with spatiotemporal dual-constraint pruning are adopted. The TW algorithm obtains a fine-grained executable trajectory that satisfies both dynamic and temporal constraints. The emergency anchor point rollback module is used for... And the sequence of primary communication anchor points, to obtain the latest decision time for each primary communication anchor point. and alternative communication anchor list A primary and secondary communication anchor point decision tree is constructed, which is the emergency decision tree for AUV execution. The output module is used to output complete planning products including mission plan, cruise path, communication plan and emergency handover strategy.

[0061] Through the description of several embodiments of the communication window-aware AUV hierarchical path planning method and system of the present invention, it can be seen that the present invention has at least one or more of the following advantages: 1) By modeling multimodal communication resources in a unified spatiotemporal communication anchor point, heterogeneous communication opportunities such as Iridium satellite overhead, underwater acoustic relay, radio line-of-sight, and Wi-Fi near-field can be jointly solved under a single planning framework. 2) By constructing a spatiotemporal extended planning graph, communication opportunities are inverted into strong spatiotemporal coupling constraints in the path planning stage, thereby realizing the spatiotemporal binding of communication events and path nodes; 3) By constructing a hierarchical optimization architecture with three objectives—task coverage completeness, energy consumption, and communication backhaul completeness—the macroscopic skeleton search and local dynamic refinement are decoupled, balancing global optimality and engineering solvability; through a spatiotemporal dual-constraint pruning strategy, in A... During the TW search process, both spatial reachability and time window reachability are checked simultaneously to ensure that the time of arrival at the communication anchor point falls strictly within the available time window, and pruning can significantly reduce the search space size. 4) By using multi-objective weight adaptive settings driven by working condition templates, the weights of task coverage integrity cost, energy consumption cost, and communication backhaul integrity cost can be automatically set according to the AUV task working conditions, which significantly reduces the operator's burden and improves planning consistency. 5) By constructing a primary and alternative communication anchor point decision tree and a latest decision time mechanism, the complete emergency switching strategy required for the online execution phase of AUV can be output during the planning phase, providing emergency rollback capability during the online execution phase, thereby improving the robustness of task execution.

[0062] Finally, it should be noted that the various embodiments in this specification are described in a progressive manner, with each embodiment focusing on the differences from other embodiments. The same or similar parts between the various embodiments can be referred to each other.

[0063] The above embodiments are only used to illustrate the technical solutions of the present invention and not to limit them; although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications can still be made to the specific implementation of the present invention or equivalent substitutions can be made to some technical features without departing from the spirit of the technical solutions of the present invention, and all such modifications and substitutions should be covered within the scope of the technical solutions claimed in the present invention.

Claims

1. A communication window-aware AUV hierarchical path planning method, characterized in that, Includes the following steps: S1. Obtain input information, including mission intent and constraints, satellite ephemeris, mother ship or base station navigation plan, ocean current and environmental forecasts, AUV parameters and operating conditions; S2. Generate a set of spatiotemporal communication anchor points based on the input information. , No. Communication anchor points Represented as equation (1); where, , , , , , and The first The planar coordinates of each communication anchor point, the depth range of the required AUV, the time window, the link type, the estimated available bandwidth, the reliability estimate, and the buoyancy risk score, wherein the link type includes satellite link, radio link, underwater acoustic relay link, and Wi-Fi link; (1)； S3. Construct a spatiotemporal planning network, including: constructing a spatial base map of the task region according to grids or topological key points. ;right Each node in the dataset is time-extended along a discretized time axis corresponding to the AUV's expected arrival time, resulting in a spatiotemporal node set. For each Generate communication event nodes and will and Connect all nodes that satisfy the triple matching of space, time, and depth to obtain a spatiotemporal extended graph. ; S4. Perform global hierarchical planning, including: Constructing a multi-objective optimization function , expressed as equation (3); where, for An effective path on the way, , and These are the costs associated with task coverage integrity, energy consumption, and communication backhaul integrity, respectively. , and These are, respectively, the task coverage integrity weight, energy consumption weight, and communication return integrity weight driven by the working condition template; (3)； To minimize With the goal of, in An improved label correction algorithm is used to search the Pareto non-dominated solution set to obtain the macroscopic path skeleton. , This includes the primary communication anchor sequence and coarse-grained arrival time windows; S5. Perform local dynamic refinement, including: For the target skeleton, an improved A algorithm with spatiotemporal dual-constraint pruning is adopted. The TW algorithm solves the sub-path segment by segment, introduces an ocean current compensation model to correct the actual arrival time of the AUV, and obtains a fine-grained executable trajectory through dual verification of spatial reachability and temporal reachability. ; S6. Output complete planning deliverables, including mission plan, cruise route, and communication plan.

2. The communication window-aware AUV hierarchical path planning method according to claim 1, characterized in that, It also includes an emergency handover strategy generation step, which includes: based on And the sequence of primary communication anchor points, setting the latest decision time for each primary communication anchor point. Within the preset time neighborhood of each primary communication anchor point, multiple alternative communication anchor points of the same and different types are searched and sorted to obtain a list of alternative communication anchor points, thereby constructing a primary + alternative communication anchor point decision tree. When the AUV is operating online, the primary communication anchor point is selected first. If it is determined before the latest decision time that the AUV cannot reach the primary communication anchor point, the alternative communication anchor points in the alternative communication anchor point list are tried in sequence until the first feasible alternative communication anchor point is hit.

3. The communication window-aware AUV hierarchical path planning method according to claim 2, characterized in that, In the emergency handover strategy generation step, the latest decision time for each primary communication anchor point. According to equation (9), where, For the first The start time of each communication anchor point time window For decision-making and switching margin, it is calculated according to equation (10). The shortest time required to determine AUV deviation. The time required for switching to an alternative communication anchor point, For safety margin; (9)； (10)； The sorting strategies for multiple candidate communication anchor points include ascending order based on the travel cost of switching from the primary communication anchor point to a candidate communication anchor point, descending order based on the reliability estimate of the candidate communication anchor points, and ascending order based on the time proximity of the candidate communication anchor points. The top-ranked candidates are then selected. An ordered linked list of alternative communication anchor points is constructed from the available alternative communication anchor points. .

4. The communication window-aware AUV hierarchical path planning method according to claim 2, characterized in that, In step S3, the spatial base map , for The set of nodes, for The set of edges; for Each node in Discretized timeline based on AUV's estimated arrival time Perform time extension to obtain a set of spatiotemporal nodes. , expressed as equation (2); where, For a moment, At the starting time, The time spent walking is long; (2)； The spatial, temporal, and depth triple matching includes: nodes With communication anchor point The distance between them does not exceed the preset distance threshold and time. Landing at the communication anchor point Time window Indoor and AUV can be used at any time Reaching depth range ; Spacetime Extension Graph , for The set of vertices, , for The set of edges, ;in, For the reason Derived spacetime edges, for Nodes that satisfy spatial, temporal, and depth matching The edges connecting them.

5. The communication window-aware AUV hierarchical path planning method according to claim 4, characterized in that, In step S4, the task coverage completeness cost Represented as equation (4), Assigning weights to each task. For the task Completion rate; energy consumption cost Represented as equation (5), AUV speed relative to water The following sailing power, for The edges in For AUV along The sailing time Additional energy consumption during AUV ascent or descent. and These are the lower and upper limits of energy consumption normalization, respectively. Communication backhaul integrity cost Represented as equation (6), Preset priority weights for communication feedback during the task phase; (4)； (5)； (6)。 6. The communication window-aware AUV hierarchical path planning method according to claim 5, characterized in that, Step S4 also includes optimizing the multi-objective function. Set energy constraints, buoyancy feasibility constraints, and weight constraints; the energy constraints are as follows: ,in, For AUV along Energy consumption during navigation Available power for AUV For battery margin, The aforementioned constraint on the feasibility of ascent is that, at any spatiotemporal node of the ascent segment, the AUV... Capable of rising to The dynamic feasibility; the weight constraint condition is , , and All are greater than 0.

7. The communication window-aware AUV hierarchical path planning method according to claim 6, characterized in that, In step S4, the operational mode template includes long-range survey, emergency search, energy-saving cruise, communication support, and multi-mission integration; when the AUV is in the long-range survey operational mode, , , When the AUV is in emergency search and rescue operation, , , 0.6; When the AUV is in energy-saving cruise mode, , , ; When the AUV is in a communication-assured working condition , , ; When the AUV is in a multi-tasking integrated operating condition , , .

8. The communication window-aware AUV hierarchical path planning method according to claim 7, characterized in that, In step S5, the correction of the actual arrival time of the AUV includes: Establish the current compensation model shown in Equation (7) to calculate the actual ground velocity of the AUV. ;in, Here are the position coordinates of the AUV. For a moment, The water velocity of the AUV, For ocean current vector field; (7)； For each edge The actual arrival time of the AUV is solved iteratively according to equation (8). ,in, for Location of the AUV at any given time For heading; (8)。 9. The communication window-aware AUV hierarchical path planning method according to claim 8, characterized in that, In step S5, the improved A with spatiotemporal dual-constraint pruning... The TW algorithm includes the following sub-steps: S51, Read in the sub-path starting point, target communication anchor point and its time window, and ocean current field; S52. Add the starting point of the sub-path to the OPEN list, set its cumulative cost to 0, and set its estimate to heuristic estimate. S53. Retrieve the OPEN state with the lowest estimated value from the OPEN list. ,in, For planar position, For the course, The time when this state is reached; determine Has the target communication anchor point been reached? If yes, backtrack the parent pointer, output the fine sub-path, and the algorithm terminates; otherwise, proceed with... Perform neighborhood expansion and retrieve the state of an unprocessed neighbor. ; S54, to Perform spatial reachability verification to determine Do they satisfy simultaneously? Located within the mission domain and by arrive Satisfy AUV dynamic constraints; otherwise, then Discard and return to sub-step S53 to continue execution; if so, continue with... Perform time window reachability verification to determine Can it arrive within the time window of the target communication anchor point? If not, then... Discard and return to sub-step S53 to continue execution; if so, calculate. Actual arrival time and transit time arrive The new cumulative cost; S55. Determine if the new cumulative cost is less than... If the recorded cost is not found, return to step S53 to continue execution; if so, update... The cumulative cost, valuation, and parent pointer, and will Add to the OPEN list and return to step S53 to continue execution; the algorithm terminates when the fine sub-paths are output or the OPEN list is exhausted.

10. A communication window-aware AUV hierarchical path planning system, characterized in that, AUV hierarchical path planning method for performing communication window awareness as described in any one of claims 2 to 9; The communication window-aware AUV hierarchical path planning system includes: The input module is used to obtain input information including mission intent and constraints, satellite ephemeris, mother ship or base station navigation plan, ocean current and environmental forecasts, AUV parameters and operating conditions; The communication resource modeling module is used to generate a set of spatiotemporal communication anchor points based on input information; The spatiotemporal planning network construction module is used to obtain a spatiotemporal extension graph based on the set of spatiotemporal communication anchor points and the task space description; The global hierarchical planning module is used to construct a multi-objective optimization function and use an improved label correction algorithm to search for Pareto non-dominated solution sets on the spatiotemporal extension graph to obtain the macroscopic path skeleton. The local refinement module is used to improve the AUV's performance by employing spatiotemporal dual-constraint pruning based on the macroscopic path framework, ocean current field, and AUV dynamic constraints. The TW algorithm obtains a fine-grained executable trajectory; The emergency anchor point rollback module is used to obtain the latest decision time of each primary communication anchor point and the alternative communication anchor point list, and construct a primary + alternative communication anchor point decision tree; The output module is used to output complete planning deliverables, including mission plans, cruise routes, communication plans, and emergency handover strategies.

Images