Unmanned ship real-time scheduling control system

Abstract

The invention belongs to the technical field of unmanned ship dispatching, and discloses an unmanned ship real-time dispatching control system. The system comprises an area construction module which pre-constructs observation areas corresponding to N unmanned ships based on environment perception data in a target observation sea area in unit time and state parameters of the N unmanned ships; the first processing module is used for performing feature extraction on the environmental perception data in the target observation sea area in unit time based on the observation areas corresponding to the N unmanned ships to obtain environmental perception feature data corresponding to the N unmanned ships; the second processing module is used for performing feature extraction based on the state parameters of the N unmanned ships in unit time to obtain state feature data corresponding to the N unmanned ships; the scheduling decision-making module is used for inputting the environment perception characteristic data and the state characteristic data corresponding to the N unmanned ships into a scheduling decision-making level setting model to obtain corresponding scheduling decision-making levels; and intelligent scheduling and cooperative control of multiple unmanned ships are realized.

Description

Technical Field

[0001] The present invention relates to the technical field of unmanned ship scheduling, and more specifically, to a real-time scheduling control system for an unmanned ship. Background Art

[0002] With the increasing development of tasks such as marine resource surveying, environmental monitoring, and disaster warning, multi-point collaborative observation based on swarms of unmanned vessels has become an important development direction for marine observation technology. Existing unmanned vessel scheduling and control systems primarily implement simple division of labor based on initial mission conditions through task pre-assignment, path planning, and static scheduling strategies.

[0003] However, in actual applications, the marine environment is highly dynamic and uncertain. External environmental parameters such as wind speed, ocean currents, and solar radiation intensity fluctuate frequently. Furthermore, unmanned vessels (UAVs) can degrade during operation due to prolonged navigation, experiencing issues such as yaw drift, rapid battery drain, and unstable attitude. Existing scheduling and control technologies generally lack the ability to dynamically perceive and respond to these environmental changes and the UAV's own operating status, making it impossible to adaptively adjust and fine-tune task scheduling strategies. This can cause some UAVs to withdraw from observations midway due to abnormal conditions during mission execution, leading to a break in the observation data chain and an imbalance in system load, impacting the overall mission continuity and observation stability of the swarm system.

[0004] Therefore, there is an urgent need to provide an intelligent scheduling and control system that integrates multi-dimensional environmental perception and operation status assessment to improve the resource allocation accuracy, adaptive scheduling capability and task execution robustness of unmanned ship clusters in complex and dynamic marine environments. Summary of the Invention

[0005] In order to overcome the above-mentioned defects of the prior art and achieve the above-mentioned purpose, the present invention provides the following technical solution: a real-time dispatching and control system for an unmanned vessel, comprising:

[0006] The area construction module pre-constructs the observation area corresponding to N unmanned ships based on the environmental perception data in the target observation sea area per unit time and the status parameters of N unmanned ships;

[0007] The first processing module extracts features of the environmental perception data within the target observation sea area per unit time based on the observation areas corresponding to the N unmanned ships, and obtains the environmental perception feature data corresponding to the N unmanned ships;

[0008] The second processing module extracts features based on the state parameters of the N unmanned ships per unit time to obtain state feature data corresponding to the N unmanned ships;

[0009] The scheduling decision module is used to input the environmental perception feature data and state feature data corresponding to N unmanned ships into the scheduling decision level setting model to obtain the corresponding scheduling decision level;

[0010] The intelligent scheduling module intelligently schedules N unmanned ships based on their corresponding environmental perception feature data, state feature data and scheduling decision levels.

[0011] Furthermore, the scheduling decision level includes a first-level scheduling level, a second-level scheduling level, and a third-level scheduling level; the method for intelligently scheduling N unmanned ships includes:

[0012] The unmanned ships corresponding to the first-level scheduling decision level are constructed into a first-level scheduling unmanned ship set;

[0013] The unmanned ships corresponding to the second-level scheduling decision level are constructed into a second-level scheduling unmanned ship set, and the unmanned ships in the second-level scheduling unmanned ship set are intelligently scheduled;

[0014] The unmanned ships corresponding to the third-level scheduling decision level are constructed into a three-level scheduling unmanned ship set;

[0015] Based on the first-level scheduling unmanned ship set and the third-level scheduling unmanned ship set, the corresponding environmental perception feature data and state feature data are analyzed and processed to obtain a set of replacement candidate triples;

[0016] The unmanned ships in the first-level scheduling unmanned ship set and the second-level scheduling unmanned ship set are intelligently scheduled based on the replacement candidate triplet set.

[0017] Furthermore, the method for intelligently scheduling the unmanned ships in the secondary scheduling unmanned ship set includes:

[0018] Obtain each unmanned ship in the secondary scheduling unmanned ship set, input the corresponding environmental perception feature data and state feature data into the unmanned ship parameter setting model, and obtain the corresponding unmanned ship target adjustment parameters. The unmanned ship target adjustment parameters include propulsion power, rudder angle adjustment frequency, and attitude control accuracy; adjust the current state parameters of the unmanned ship based on the unmanned ship target adjustment parameters to upgrade the unmanned ship from the secondary scheduling level to the tertiary scheduling level; the details are as follows:

[0019] S700: Record the number of unmanned ships in the secondary scheduling unmanned ship set as EJ; let the initial value of ej be 1, and the value range of ej be 1 to EJ;

[0020] S701: Obtain the ejth unmanned ship from the secondary scheduling unmanned ship set, input the environmental perception feature data and state feature data of the ejth unmanned ship into the unmanned ship parameter setting model, and obtain the unmanned ship adjustment parameters corresponding to the ejth unmanned ship, which are recorded as the unmanned ship target adjustment parameters; the unmanned ship adjustment parameters include propulsion power, rudder angle adjustment frequency, and attitude control accuracy;

[0021] S702: Adjust the current state parameters of the ejth unmanned ship to the unmanned ship target adjustment parameters; so that the unmanned ship is upgraded from the second-level scheduling level to the third-level scheduling level;

[0022] S703: Let ej=ej+1. If ej is less than or equal to EJ, continue executing S701 to S702; if ej is greater than EJ, end the current process.

[0023] Furthermore, the method for obtaining the replacement candidate triple set includes:

[0024] The environmental perception feature data and state feature data corresponding to the first-level dispatched unmanned ship and the third-level dispatched unmanned ship are constructed as the replacement fitness data, and the replacement fitness between each group of first-level and third-level unmanned ships is calculated based on the data;

[0025] According to the succession fitness result, a triple containing the serial number of the unmanned ship to be replaced, the serial number of the candidate succession unmanned ship and the corresponding fitness is constructed to form a succession candidate triple set; the details are as follows:

[0026] S500: The number of unmanned ships in the first-level scheduling unmanned ship set is recorded as YJ, and the number of unmanned ships in the third-level scheduling unmanned ship set is recorded as SJ; let the initial value of yj be 1, and the value range of yj is 1 to YJ; let the initial value of sj be 1, and the value range of sj is 1 to SJ;

[0027] S501: constructing the environmental perception feature data and state feature data of the yj-th unmanned ship in the first-level scheduling unmanned ship set into the first replacement fitness data; constructing the environmental perception feature data and state feature data of the sj-th unmanned ship in the third-level scheduling unmanned ship set into the second replacement fitness data;

[0028] S502: Calculate the succession fitness between the yj-th unmanned ship and the sj-th unmanned ship based on the first succession fitness data and the second succession fitness data; construct a succession candidate triplet by using yj, sj and the succession fitness data, and add the succession candidate triplet to the succession candidate triplet set;

[0029] S503: Let sj = sj + 1. If sj is less than or equal to SJ, continue executing S501 to S502. If sj is greater than SJ, let yj = yj + 1. If yj is less than or equal to YJ, let sj = 1 and continue executing S501 to S502. If yj is greater than YJ, obtain the successor candidate triple set and end the current process.

[0030] Furthermore, the method for intelligently scheduling the unmanned ships in the first-level scheduling unmanned ship set and the second-level scheduling unmanned ship set based on the replacement candidate triplet set includes:

[0031] Obtain the candidate triplet with the highest succession fitness. Based on the observation area of ​​the unmanned ship to be replaced and the candidate unmanned ship contained in the candidate triplet, recalculate the updated observation area of ​​the two according to the preset observation area succession ratio, and use them as the scheduling optimization results.

[0032] Remove all candidates involved in the used successor candidate triples, and repeat the above operation for the remaining successor candidate triples until the successor candidate triple set is empty; the details are as follows:

[0033] S600: Obtaining the succession candidate triplet with the highest succession fitness from the succession candidate triplet set, and recording it as the current succession candidate triplet; the first digit of the succession candidate triplet is the serial number of the unmanned ship to be replaced; the second digit of the succession candidate triplet is the serial number of the candidate succession unmanned ship; the third digit of the succession candidate triplet is the succession fitness when the unmanned ship corresponding to the second digit replaces the unmanned ship corresponding to the first digit;

[0034] S601: Obtain the observation area of ​​the unmanned ship corresponding to the first position of the candidate triplet, which is recorded as the first observation area; obtain the observation area of ​​the unmanned ship corresponding to the second position of the candidate triplet, which is recorded as the second observation area;

[0035] S602: Multiplying the area of ​​the first observation area by a preset observation area replacement ratio to obtain an apportioned observation area area; subtracting the first observation area from the apportioned observation area to obtain a first updated observation area area, where the first updated observation area area is the updated observation area area of ​​the unmanned ship corresponding to the first position of the candidate triplet; summing the second observation area area and the apportioned observation area area to obtain a second updated observation area area, where the second updated observation area area is the updated observation area area of ​​the unmanned ship corresponding to the second position of the candidate triplet;

[0036] S603: Remove all successor candidate triples corresponding to the first or second position in the current successor candidate triple from the successor candidate triple set;

[0037] S604: Repeat S600 to S603, and stop executing when the set of successor candidate triples is empty.

[0038] Furthermore, the method of pre-constructing the observation areas corresponding to N unmanned ships includes:

[0039] Acquire environmental perception data within the target observation sea area, the environmental perception data including ocean current velocity field, ocean current direction map, wind speed, wind direction, wave level and solar radiation intensity;

[0040] Obtaining status parameters of N unmanned ships, including real-time position coordinates, battery power, navigation speed, and heading;

[0041] The state parameters and environmental perception data corresponding to N unmanned ships are input into the comprehensive feasibility evaluation model respectively to obtain the comprehensive feasibility scores corresponding to the N unmanned ships;

[0042] The comprehensive feasibility scores corresponding to the N unmanned ships are normalized to obtain the regional allocation factors corresponding to the N unmanned ships. Combined with the total area of ​​the target observation sea area, the observation area that the N unmanned ships should undertake is calculated; the observation area that the N unmanned ships should undertake is the observation area corresponding to the N unmanned ships.

[0043] Furthermore, the method for acquiring the environmental perception feature data corresponding to the N unmanned ships includes:

[0044] Obtain the environmental perception data of the observation area corresponding to each unmanned vessel, calculate the mean current velocity and the current velocity stability index of the current velocity field, and calculate the mean current direction and the current direction stability index of the current direction map;

[0045] Calculate the solar energy availability index based on the solar radiation intensity, obtain the wind speed disturbance degree and wind direction jump frequency index based on the wind speed and wind direction data, and obtain the wind energy disturbance level based on the preset wind energy disturbance level mapping table;

[0046] The indicators calculated above are constructed into the environmental perception feature data of each unmanned ship; the details are as follows:

[0047] S100: Let n be initialized to 1, and the value range of n is 1 to N; divide the unit time into G time points;

[0048] S101: Obtain environmental perception data corresponding to the observation area of ​​the nth unmanned vessel; calculate the corresponding current velocity mean and current velocity stability index based on the current velocity field in the environmental perception data; calculate the corresponding current direction mean and current direction stability index based on the current direction map in the environmental perception data; calculate the corresponding solar energy availability index based on the solar radiation intensity in the environmental perception data;

[0049] The corresponding wind speed disturbance degree is calculated based on the wind speed in the environmental sensing data; the corresponding wind direction jump frequency index is calculated based on the wind direction in the environmental sensing data; the corresponding wind energy disturbance level index is calculated based on the wind speed disturbance degree and the wind direction jump frequency index; the wind energy disturbance level index is matched with the pre-built wind energy disturbance level mapping table to obtain the corresponding wind energy disturbance level;

[0050] S102: constructing the environmental perception feature data of the nth unmanned vessel using the mean ocean current speed, the ocean current speed stability index, the mean ocean current direction, the ocean current direction stability index, the solar energy availability index, and the wind energy disturbance level;

[0051] S103: Let n=n+1. If n is less than or equal to N, continue executing S101 to S102. If n is greater than N, obtain the environmental perception feature data corresponding to N unmanned ships and end the current process.

[0052] Furthermore, the method for obtaining the state characteristic data corresponding to the N unmanned ships includes:

[0053] Obtain the real-time status parameters of each unmanned vessel, identify the deviation mark of the unmanned vessel by analyzing the change trend of the position coordinates, analyze the change trend of the battery power to obtain the type of power change trend, and calculate the navigation stability index based on the navigation speed and heading stability;

[0054] The unmanned ship deviation mark, power change trend type and navigation stability index are constructed into the state characteristic data corresponding to each unmanned ship; the details are as follows:

[0055] S200: Let n be initially set to 1, and the value range of n is 1 to N; divide the unit time into G time points;

[0056] S201: Acquire state parameters of the nth unmanned ship; analyze and process the position coordinate change trend based on the real-time position coordinates in the state parameters to obtain an unmanned ship deviation indicator of the nth unmanned ship; analyze and process the battery power change trend based on the battery power in the state parameters to obtain the power change trend type of the nth unmanned ship; calculate the navigation stability index of the nth unmanned ship based on the navigation speed and heading in the state parameters;

[0057] S202: Constructing the state characteristic data of the nth unmanned ship by using the unmanned ship deviation mark, the power change trend type, and the navigation stability index;

[0058] S203: Let n=n+1. If n is less than or equal to N, continue to execute S201 to S202; if n is greater than N, obtain the state feature data corresponding to N unmanned ships and end the current process.

[0059] Furthermore, the method for obtaining the unmanned ship deviation mark of the nth unmanned ship includes:

[0060] The closest distance between the historical track point of the unmanned ship and the boundary of the observation area is calculated, and the proportion of track points that cross the boundary is counted; if the proportion exceeds the preset threshold of the out-of-bounds ratio, the unmanned ship deviation flag is set to yes, otherwise it is set to no; the details are as follows:

[0061] S300: Obtain the observation area corresponding to the nth unmanned ship and extract the boundary coordinates corresponding to the observation area; set the initial value of g to 1, and the value range of g is 1 to G; set the initial value of the counting variable that exceeds the boundary to 0;

[0062] S301: Obtain the real-time position coordinates of the nth unmanned ship at the gth time point, calculate the Euclidean distance between the real-time position coordinates and all boundary coordinates of the observation area, and select the smallest Euclidean distance as the unmanned ship boundary distance corresponding to the gth time point. The unmanned ship boundary distance refers to the closest distance between the unmanned ship and the boundary of the observation area.

[0063] S302: If the unmanned ship boundary distance corresponding to the g-th time point is greater than or equal to the preset boundary safety distance threshold, the boundary exceeding count variable is incremented by one;

[0064] S303: Let g = g + 1. If g is less than or equal to G, continue executing S301 to S302; if g is greater than G, execute S304.

[0065] S304: Divide the out-of-bounds count variable by G to obtain the out-of-bounds ratio. If the out-of-bounds ratio is greater than or equal to the preset out-of-bounds ratio threshold, the unmanned ship deviation flag is set to yes; if the out-of-bounds ratio is less than the preset out-of-bounds ratio threshold, the unmanned ship deviation flag is set to no.

[0066] Furthermore, the method for obtaining the power change trend type of the nth unmanned ship includes:

[0067] Calculate the battery charge change rate based on the battery charge at consecutive time points. Combined with the preset battery charge change rate threshold 1 and battery charge change rate threshold 2, count the number of stable charge changes, slow charge decreases, and rapid charge decreases respectively.

[0068] The change type with the most changes is taken as the power change trend type of the nth unmanned ship; the details are as follows:

[0069] S400: Assume that the initial value of t is 1, and the value range of t is 1 to G-1; preset a battery power change rate threshold 1 and a battery power change rate threshold 2, wherein the battery power change rate threshold 1 is less than the battery power change rate threshold 2; and set the initial values ​​of the stable power change number, the slowly decreasing power change number, and the rapidly decreasing power change number to 0;

[0070] S401: Obtain the battery power of the nth unmanned ship at the tth time point; obtain the battery power of the nth unmanned ship at the t+1th time point; calculate the battery power change rate between the tth and t+1th time points based on the battery power at the tth and t+1th time points;

[0071] S402: If the battery charge change rate is less than the battery charge change rate threshold 1, then the number of stable charge changes is increased by one; if the battery charge change rate is greater than or equal to the battery charge change rate threshold 1 and less than the battery charge change rate threshold 2, then the number of slowly decreasing charge changes is increased by one; if the battery charge change rate is greater than or equal to the battery charge change rate threshold 2, then the number of rapidly decreasing charge changes is increased by one;

[0072] S403: Let t = t + 1. If t is less than or equal to G - 1, continue executing S401 to S402; if t is greater than G - 1, execute S404.

[0073] S404: If the number of stable changes in power is the largest, the power change trend type is the stable change type; if the number of slowly decreasing changes in power is the largest, the power change trend type is the slowly decreasing change type; if the number of rapidly decreasing changes in power is the largest, the power change trend type is the rapidly decreasing change type.

[0074] Compared with the existing technology, the technical effects and advantages of the unmanned ship real-time scheduling control system of the present invention are as follows:

[0075] The present invention provides a real-time dispatching and control system for unmanned vessels, which constructs an intelligent dispatching mechanism that integrates environmental perception characteristics and state characteristics, and can realize intelligent task allocation and coordinated control of multiple unmanned vessels in a dynamic ocean environment.

[0076] By deploying edge computing nodes, the system performs distributed collection and real-time analysis of environmental perception data, including ocean current velocity fields, current patterns, wind speed, wind direction, wave levels, and solar radiation intensity in the target observation area. Combined with the unmanned vessel's real-time location, battery charge, speed, and heading, the system generates environmental perception and status characteristic data for each unmanned vessel. The system further inputs this characteristic data into a comprehensive feasibility assessment model constructed using a multi-layer perceptron neural network to obtain a comprehensive feasibility score for each unmanned vessel, thereby enabling dynamic observation area division based on capability differences.

[0077] On this basis, through the scheduling decision level setting model, the operating status of the unmanned ship is intelligently divided into multiple levels to support differentiated scheduling response strategies: for unmanned ships of the first-level scheduling level, by constructing a set of succession candidate triples and calculating the succession fitness, automatic task transfer and relay scheduling are realized; for unmanned ships of the second-level scheduling level, targeted control adjustment parameters are generated based on the pre-trained machine learning model and sent to execution units such as propulsion power, rudder angle control, and attitude stability control, so as to promote the operating status of the unmanned ship to be stable and have the ability to switch to the third-level scheduling level. Through the above steps, the present invention realizes a closed-loop intelligent scheduling mechanism of unmanned ship state perception, scheduling level determination, task load adjustment, and state optimization feedback, which significantly improves the system's resource allocation accuracy, adaptive scheduling capability, and task execution robustness in complex and unstructured marine environments. BRIEF DESCRIPTION OF THE DRAWINGS

[0078] Figure 1 This is a schematic diagram of a real-time dispatching and control system for an unmanned vessel according to embodiment 1 of the present invention;

[0079] Figure 2 This is a flow chart of a real-time dispatching control method for an unmanned ship according to embodiment 3 of the present invention;

[0080] Figure 3 This is a schematic diagram of a real-time dispatching and control system for an unmanned vessel according to embodiment 2 of the present invention;

[0081] Figure 4 Flowchart of a method for intelligently dispatching N unmanned ships;

[0082] Figure 5 The present invention is a flow chart of a method for intelligently scheduling unmanned ships in a first-level scheduling unmanned ship set and a second-level scheduling unmanned ship set based on a set of succession candidate triples. DETAILED DESCRIPTION

[0083] The technical solutions in the embodiments of the present invention will be described in detail, clearly and completely below in conjunction with the accompanying drawings in the embodiments of the present invention. It should be noted that the specific embodiments described below are only used to better illustrate and describe the technical solutions of the present invention, and are intended to enable those skilled in the art to better understand and implement the present invention, and should not be construed as limiting the scope of protection of the present invention. Without departing from the spirit and essence of the present invention, those skilled in the art may modify, adjust or make equivalent replacements based on the contents disclosed in the present invention, and these should all be regarded as the scope of protection of the present invention.

[0084] Example 1

[0085] See also Figure 1As shown, this embodiment discloses a real-time scheduling and control system for an unmanned ship, including a regional construction module, a first processing module, a second processing module, a scheduling decision module and an intelligent scheduling module. Each module realizes data transmission through wired and / or wireless connections.

[0086] The area construction module pre-constructs the observation area corresponding to N unmanned ships based on the environmental perception data in the target observation sea area and the status parameters of N unmanned ships within a unit time.

[0087] The method of pre-constructing the observation area corresponding to N unmanned ships includes:

[0088] Acquire environmental perception data within the target observation sea area, the environmental perception data including ocean current velocity field, ocean current direction map, wind speed, wind direction, wave level and solar radiation intensity;

[0089] It should be noted that the ocean current velocity field and ocean current direction map can be perceived in real time by buoys and ADCP (Acoustic Doppler Current Profiler) equipment deployed in the target observation sea area. Among them, the ocean current velocity field is the vector field data describing the flow velocity of seawater at different geographical coordinates in the target sea area, and the ocean current direction map is the angular distribution map describing the flow direction of ocean currents in various regions. The wind speed, wind direction, and wave level are obtained through marine meteorological stations. The solar radiation intensity is measured in real time by the light sensor integrated on the top of the unmanned boat. The wind direction is expressed in degrees, and the unit is angle. The value range of the wind direction is 0° to 360°, which is used to indicate the direction angle of the wind direction, where 0° represents due north, 90° represents due east, 180° represents due south, 270° represents due west, and the remaining angles are similar.

[0090] Obtaining status parameters of N unmanned ships, including real-time position coordinates, battery power, navigation speed, and heading;

[0091] It should be noted that the real-time location coordinates are obtained via the Global Positioning System or Beidou navigation system, the battery charge level is obtained via the unmanned vessel's battery management system, and the navigation speed and heading are obtained via the unmanned vessel's navigation control system. The heading is expressed in degrees; the heading value ranges from 0° to 360°, indicating the heading angle, where 0° represents due north, 90° represents due east, 180° represents due south, 270° represents due west, and so on for the remaining angles.

[0092] The state parameters and environmental perception data corresponding to N unmanned ships are respectively input into the comprehensive feasibility evaluation model to obtain the comprehensive feasibility scores corresponding to the N unmanned ships; the higher the comprehensive feasibility score, the stronger the observation capability of the unmanned ship in the current environment, the more tasks it can undertake, and the larger the observation area can be divided.

[0093] The comprehensive feasibility scores corresponding to the N unmanned ships are normalized to obtain the regional allocation factors corresponding to the N unmanned ships. Combined with the total area of ​​the target observation sea area, the observation area that the N unmanned ships should undertake is calculated; the observation area that the N unmanned ships should undertake is the observation area corresponding to the N unmanned ships.

[0094] The method for obtaining the area allocation factor includes:

[0095]

[0096] Among them, FPYZ n is the regional allocation factor of the nth unmanned ship, ZHPF n is the comprehensive feasibility score of the nth unmanned ship, is the sum of the comprehensive feasibility scores of N unmanned ships, i is the index variable of the summation formula, ZHPF i is the comprehensive feasibility score of the i-th unmanned ship.

[0097] The method for obtaining the area of ​​the observation region includes:

[0098] GCMJ n =FPYZ n ×ZMJ;

[0099] Among them, GCMJ n is the observation area of ​​the nth unmanned ship, and ZMJ is the total area of ​​the target observation sea area.

[0100] It should be noted that in one embodiment of the present invention, in order to achieve efficient coverage of the target observation sea area and reasonable distribution of unmanned ship mission loads, the system proposes a method for dividing the observation area based on a comprehensive feasibility score. Specifically, the system first calculates a comprehensive feasibility score for each of the N unmanned ships to be scheduled. The score reflects the mission execution capability and stability of each unmanned ship under the current environmental conditions and its own operating status. The higher the comprehensive feasibility score, the more suitable the unmanned ship is for better environmental adaptability, more stable navigation status, and more sustainable energy support capabilities, making it suitable for undertaking larger observation mission loads.

[0101] To dynamically divide mission areas based on capability differences, the system normalizes the comprehensive feasibility scores of all unmanned vessels and constructs a regional allocation factor for the nth unmanned vessel. This factor represents the proportion of the unmanned vessel's score within the cluster, and therefore its weight in the overall observation mission. A larger regional allocation factor indicates a more capable unmanned vessel, making it suitable for larger observation areas. Based on this regional allocation factor, the system further multiplies it by the total area of ​​the target observation sea area to determine the observation area corresponding to the nth unmanned vessel.

[0102] Through the above derivation process, the feasibility score obtained by fusing multi-source features is converted into a spatial division basis for observation mission areas. This allows for dynamic allocation of observation mission areas based on the comprehensive capabilities of each unmanned vessel, ensuring more reasonable, flexible, and stable task distribution and resource utilization in the system. This method not only avoids the resource waste or overload risks that may arise from static average allocation, but also enhances the system's real-time scheduling capabilities and task completion rate in complex and dynamic environments.

[0103] The training method of the comprehensive feasibility assessment model includes:

[0104] Pre-constructing a comprehensive feasibility assessment dataset, wherein the comprehensive feasibility assessment dataset includes Y groups of comprehensive feasibility assessment data and comprehensive feasibility scores corresponding to the Y groups of comprehensive feasibility assessment data, where Y is a positive integer greater than 0, and the comprehensive feasibility assessment data includes environmental perception data and state parameters of the unmanned vessel; dividing the comprehensive feasibility assessment dataset into a comprehensive feasibility assessment data training set and a comprehensive feasibility assessment data validation set, wherein the comprehensive feasibility assessment data training set is used for parameter learning of the comprehensive feasibility assessment model, and the comprehensive feasibility assessment data validation set is used for real-time evaluation of the generalization ability of the comprehensive feasibility assessment model;

[0105] During the training process of the comprehensive feasibility evaluation model, a deep neural network structure based on multi-layer perceptron is adopted to convert the comprehensive feasibility evaluation data into a feature vector as input, extract the nonlinear features in the data through the hidden layer, and finally use the softmax activation function in the output layer to generate the probability distribution of the comprehensive feasibility score, and output the comprehensive feasibility score corresponding to the maximum probability as the final prediction result; the training process aims to minimize the cross-entropy loss function, and at the same time introduces an early stopping strategy to monitor the performance of the comprehensive feasibility evaluation data validation set. When the prediction accuracy on the comprehensive feasibility evaluation data validation set reaches the preset accuracy, the comprehensive feasibility evaluation model is considered to have converged and the training stops immediately.

[0106] It should be noted that when training the comprehensive feasibility assessment model, the environmental perception data includes the ocean current velocity field, ocean current direction map, wind speed, wind direction, wave level and solar radiation intensity, and the state parameters of the unmanned ship include real-time position coordinates, battery power, navigation speed and heading; the environmental perception data and the state parameters of the unmanned ship are normalized and quantified into parameters that can be used for machine learning model training. Specifically, for example, in order to achieve unified modeling and quantifiable processing of multi-dimensional heterogeneous data, the system normalizes various types of raw data and converts them into a dimensionally consistent, dimensionless standardized parameter feature set. The quantized data has a value range of [0, 1]. When pre-constructing the comprehensive feasibility assessment data set, the comprehensive feasibility score can be comprehensively evaluated by those skilled in the art based on the normalized data to obtain a comprehensive feasibility score corresponding to each set of comprehensive feasibility assessment data.

[0107] The range of the comprehensive feasibility score is also in [0, 1]. For example, if the score value is close to 1, it means that the current state of the unmanned ship is stable, the environmental adaptability is strong, and it has the priority ability to perform the main observation task; if the score value is in the middle range (such as 0.4-0.7), it means that its operating state has slight fluctuations or resources are limited, and the task load needs to be appropriately reduced; if the score value is lower than 0.4, it indicates that its operating capability is significantly weakened or the risk of failure is high, and it is not suitable to continue to undertake the original task, and the relay or evacuation strategy should be triggered.

[0108] It should be noted that environmental perception data can comprehensively reflect the dynamic environmental characteristics of the target observation sea area, among which the current speed and direction affect the navigation stability and energy consumption of the unmanned ship, the wind speed and direction affect the navigation deviation and propulsion efficiency, the wave level reflects the anti-interference ability requirements of sea surface operation, and the solar radiation intensity affects the effectiveness of solar power supply; the status parameters of N unmanned ships can reflect the current operating capability, energy status and task scheduling response capability of each unmanned ship.

[0109] By combining environmental perception data and the state parameters of the unmanned vessels with a comprehensive feasibility assessment model, each unmanned vessel is quantitatively scored for its feasibility of carrying out its observation mission under current conditions. The resulting comprehensive feasibility score reflects the vessel's overall capability to carry out the mission in the target sea area. Based on this comprehensive feasibility score, an observation area is then allocated to each unmanned vessel, matching its capabilities to achieve dynamic balancing of mission load and optimal resource utilization.

[0110] The division of the observation area not only fully considers the environmental change characteristics of the target sea area and the dynamic performance differences of the unmanned ships, but also forms an exclusive and clearly defined operating range for each unmanned ship, which helps the subsequent path planning process to generate the optimal path within the limited area, reduce path conflicts and redundant intersections, and improve the accuracy of track control and the convergence efficiency of the scheduling algorithm. In addition, based on the divided observation area, individualized path planning can be further combined with key parameters such as the remaining power, sensor status, and heading attitude of each unmanned ship to implement an adaptive control strategy for individual status, so that the task execution is more in line with the actual capability boundary, and the success rate of task completion and energy utilization efficiency are improved. Overall, a multi-level collaborative optimization effect from "regional load division" to "individual path optimization" is achieved.

[0111] The first processing module extracts features of the environmental perception data within the target observation sea area per unit time based on the observation areas corresponding to the N unmanned ships, and obtains the environmental perception feature data corresponding to the N unmanned ships;

[0112] The method for obtaining the environmental perception feature data corresponding to N unmanned ships includes:

[0113] S100: Let n be initialized to 1, and the value range of n is 1 to N; divide the unit time into G time points;

[0114] S101: Obtain environmental perception data corresponding to the observation area of ​​the nth unmanned vessel; calculate the corresponding current velocity mean and current velocity stability index based on the current velocity field in the environmental perception data; calculate the corresponding current direction mean and current direction stability index based on the current direction map in the environmental perception data; calculate the corresponding solar energy availability index based on the solar radiation intensity in the environmental perception data;

[0115] The corresponding wind speed disturbance degree is calculated based on the wind speed in the environmental sensing data; the corresponding wind direction jump frequency index is calculated based on the wind direction in the environmental sensing data; the corresponding wind energy disturbance level index is calculated based on the wind speed disturbance degree and the wind direction jump frequency index; the wind energy disturbance level index is matched with the pre-built wind energy disturbance level mapping table to obtain the corresponding wind energy disturbance level;

[0116] S102: constructing the environmental perception feature data of the nth unmanned vessel using the mean ocean current speed, the ocean current speed stability index, the mean ocean current direction, the ocean current direction stability index, the solar energy availability index, and the wind energy disturbance level;

[0117] S103: Let n=n+1. If n is less than or equal to N, continue executing S101 to S102. If n is greater than N, obtain the environmental perception feature data corresponding to N unmanned ships and end the current process.

[0118] The method for calculating the mean ocean current velocity includes:

[0119]

[0120] in, is the mean value of the ocean current velocity, g is the index variable of the summation formula, and the value range of g is 1 to G. g is the ocean current velocity corresponding to the g-th time point.

[0121] The calculation method of the ocean current velocity stability index includes:

[0122]

[0123] Among them, HLWD is the ocean current velocity stability index.

[0124] The method for calculating the mean value of the ocean current direction includes:

[0125]

[0126] in, is the mean value of the current direction, HLFX g is the current direction corresponding to the g-th time point.

[0127] The calculation method of the ocean current direction stability index includes:

[0128]

[0129] Among them, FXWD is the ocean current direction stability index.

[0130] The method for calculating the wind speed disturbance degree includes:

[0131]

[0132] Among them, FSRD is the wind speed disturbance degree, FS g is the wind speed corresponding to the g-th time point, Indicates the mean wind speed.

[0133] The calculation method of the wind direction jump frequency index includes:

[0134]

[0135] Among them, TBPL is the wind direction jump frequency index, θ thresh is the transition threshold (for example, the transition threshold can be set to 15°), FX g is the wind direction corresponding to the g-th time point, FX g-1 is the wind direction corresponding to the g-1th time point, and δ( ) is the logical judgment function.

[0136] If satisfied |FX g -FX g-1 |>θ thresh , then δ(|FX g -FX g-1 |>θ thresh ) corresponds to a value of 1. If the |FX g -FX g-1 |>θ thresh , then δ(|FX g -FX g-1 |>θ thresh ) corresponds to a value of 0.

[0137] The calculation method of the wind energy disturbance level index includes:

[0138] WDI = α × FSRD + β × TBPL;

[0139] Among them, WDI is the wind energy disturbance level index, α is the weighting coefficient of the wind speed disturbance degree, and β is the weighting coefficient of the wind direction jump frequency index. α and β reflect the importance of the wind speed disturbance degree and wind direction jump frequency index to the wind energy disturbance level index. For example, α can be set to 0.7 and β can be set to 0.3.

[0140] It should be noted that to effectively assess the degree of wind disturbance within the target observation area and improve the path planning and energy management capabilities of the unmanned vessel dispatching system under complex meteorological conditions, a wind disturbance level index based on wind speed variation and wind direction instability was designed. This index quantifies the intensity of wind disturbances per unit time and provides a key basis for unmanned vessel mission allocation, path selection, and energy budgeting.

[0141] In this application, to accurately characterize the hydrodynamic and wind energy disturbance characteristics of the target observation area per unit time, a number of mathematical indicators were constructed to characterize the stability and changing trends of the environmental state. These indicators include mean current velocity, current velocity stability index, mean current direction, current direction stability index, wind speed disturbance degree, and wind direction jump frequency. These indicators all use time series as input and extract environmental dynamic characteristics that are practical for unmanned vessel observation and scheduling tasks through statistical or rate-of-change calculations.

[0142] The mean current velocity measures the average current velocity per unit time within the unmanned vessel's observation area, reflecting the impact of current intensity on the vessel's dynamic compensation and propulsion strategy. The current velocity stability index, constructed based on the standard deviation, describes the degree of fluctuation in current velocity over time. A higher index indicates more severe current disturbances, and the greater the pressure on vessel stability and energy regulation. Similarly, the mean current direction indicates the direction of the prevailing current, helping to determine vessel heading control and predict drift risk. The current direction stability index reflects the degree of directional change and can be used to identify vortices, crosscurrents, or other complex flow disturbances, thereby assessing the risk level of the unmanned vessel's attitude control.

[0143] Furthermore, the wind speed disturbance degree, which characterizes the intensity of wind speed disturbances during a mission cycle by calculating the standard deviation of wind speed, is an important quantitative indicator for evaluating aerodynamic interference intensity. The wind direction jump frequency indicator constructs a jump determination function based on the adjacent differences in the wind direction time series. This function calculates the ratio of the number of jumps to the unit duration to quantify the sudden change in the wind field and effectively identifies whether the current situation is in a complex wind direction fluctuation zone.

[0144] The above-mentioned multiple indicators are important components of the environmental perception feature data in this application. On the one hand, they provide a highly reliable input basis for the scheduling decision level setting model. On the other hand, they can also be used as disturbance constraint input in the subsequent unmanned ship parameter adjustment model, which helps to improve the accuracy of the scheduling response and the overall anti-disturbance capability of the system.

[0145] An example of a wind energy disturbance level mapping table is shown in Table 1:

[0146] Table 1 Wind energy disturbance level mapping table

[0147] Wind Disturbance Index (WDI) Wind disturbance level 0≤WDI<0.5 1 (very weak) 0.5≤WDI<1.0 2 (weak) 1.0≤WDI<1.5 3 (medium) 1.5≤WDI<2.0 4 (strong) WDI≥2.0 5 (strong disturbance)

[0148] The method for obtaining the solar energy availability index includes:

[0149]

[0150] Among them, TYHQ is the solar energy availability index, FSQD is the solar radiation intensity, NJD is curr For current sea visibility, NJD ideal For ideal visibility at sea, SD curr is the current sea humidity, YL curr is the current sea cloud cover, ω1, ω2 and ω3 are the corresponding weighting coefficients, satisfying ω1+ω2+ω3=1. For example, ω1 can be set to 0.4, ω2 can be set to 0.3, and ω3 can be set to 0.3.

[0151] is the visibility correction term, which indicates the shading intensity of the visual environment. The lower the visibility, the stronger the shading ability of suspended particles and water vapor in the air, and the lower the sunlight transmittance. The visibility correction term is multiplied by the weighting coefficient ω1 to reflect the contribution of ω1 to the overall shading. is the humidity correction term. As humidity increases, water vapor condensation, atomization and scattering effects increase, which suppresses the radiation flux. Therefore, SD curr After normalization, multiply by the weighting coefficient ω2 as the second attenuation factor; ω3×YL curr It is the cloud cover correction item. Clouds are the most significant factor in directly blocking sunlight. The larger the cloud cover, the more serious the direct radiation is blocked.

[0152] The visibility correction, humidity correction, and cloud cover correction are combined into a total obstruction impact value. The effective solar energy scaling factor is expressed as 1-total obstruction impact value. Finally, the effective solar energy scaling factor is multiplied by the FSQD to obtain the solar energy availability index. This index not only comprehensively reflects the multi-dimensional interference of various environmental factors on photovoltaic energy supply, but also exhibits excellent flexibility and real-time performance, providing reliable lighting environment input support for unmanned vessels' path selection, power budgeting, and return determination within the scheduling cycle.

[0153] The second processing module extracts features based on the state parameters of the N unmanned ships per unit time to obtain state feature data corresponding to the N unmanned ships; the state feature data is used to assist in determining whether it is necessary to reallocate tasks or optimize path scheduling for the unmanned ships.

[0154] The method for obtaining the state characteristic data corresponding to N unmanned ships includes:

[0155] S200: Let n be initially set to 1, and the value range of n is 1 to N; divide the unit time into G time points;

[0156] S201: Acquire state parameters of the nth unmanned ship; analyze and process the position coordinate change trend based on the real-time position coordinates in the state parameters to obtain an unmanned ship deviation indicator of the nth unmanned ship; analyze and process the battery power change trend based on the battery power in the state parameters to obtain the power change trend type of the nth unmanned ship; calculate the navigation stability index of the nth unmanned ship based on the navigation speed and heading in the state parameters;

[0157] S202: Constructing the state characteristic data of the nth unmanned ship by using the unmanned ship deviation mark, the power change trend type, and the navigation stability index;

[0158] S203: Let n=n+1. If n is less than or equal to N, continue to execute S201 to S202; if n is greater than N, obtain the state feature data corresponding to N unmanned ships and end the current process.

[0159] The method for obtaining the unmanned ship deviation mark of the nth unmanned ship includes:

[0160] S300: Obtain the observation area corresponding to the nth unmanned ship and extract the boundary coordinates corresponding to the observation area; set the initial value of g to 1, and the value range of g is 1 to G; set the initial value of the counting variable that exceeds the boundary to 0;

[0161] S301: Obtain the real-time position coordinates of the nth unmanned ship at the gth time point, calculate the Euclidean distance between the real-time position coordinates and all boundary coordinates of the observation area, and select the smallest Euclidean distance as the unmanned ship boundary distance corresponding to the gth time point. The unmanned ship boundary distance refers to the closest distance between the unmanned ship and the boundary of the observation area.

[0162] S302: If the unmanned ship boundary distance corresponding to the g-th time point is greater than or equal to the preset boundary safety distance threshold, the boundary exceeding count variable is incremented by one;

[0163] S303: Let g = g + 1. If g is less than or equal to G, continue executing S301 to S302; if g is greater than G, execute S304.

[0164] S304: Divide the out-of-bounds count variable by G to obtain the out-of-bounds ratio. If the out-of-bounds ratio is greater than or equal to the preset out-of-bounds ratio threshold, the unmanned ship deviation flag is set to yes; if the out-of-bounds ratio is less than the preset out-of-bounds ratio threshold, the unmanned ship deviation flag is set to no.

[0165] It should be noted that the boundary safety distance threshold is used to determine whether the unmanned vessel is within a safe distance of the observation area, that is, the distance between the unmanned vessel and the boundary of the observation area needs to be less than the boundary safety distance threshold. The boundary safety distance threshold can be set by those skilled in the art based on the area of ​​the observation area and the volume of the unmanned vessel, for example, the boundary safety distance threshold can be set to 20 meters; the boundary excess ratio threshold can be set by those skilled in the art, for example, the boundary excess ratio threshold can be set to 0.5. By constructing an unmanned vessel deviation indicator, that is, whether the unmanned vessel approaches or exceeds the boundary of the observation area for a long time per unit time, it is possible to quickly identify whether the unmanned vessel has a tendency to deviate from the boundary of the observation area, thereby enabling early risk intervention.

[0166] The method for obtaining the power change trend type of the nth unmanned ship includes:

[0167] S400: Assume that the initial value of t is 1, and the value range of t is 1 to G-1; preset a battery power change rate threshold 1 and a battery power change rate threshold 2, wherein the battery power change rate threshold 1 is less than the battery power change rate threshold 2; and set the initial values ​​of the stable power change number, the slowly decreasing power change number, and the rapidly decreasing power change number to 0;

[0168] S401: Obtain the battery power of the nth unmanned ship at the tth time point; obtain the battery power of the nth unmanned ship at the t+1th time point; calculate the battery power change rate between the tth and t+1th time points based on the battery power at the tth and t+1th time points;

[0169] S402: If the battery charge change rate is less than the battery charge change rate threshold 1, then the number of stable charge changes is increased by one; if the battery charge change rate is greater than or equal to the battery charge change rate threshold 1 and less than the battery charge change rate threshold 2, then the number of slowly decreasing charge changes is increased by one; if the battery charge change rate is greater than or equal to the battery charge change rate threshold 2, then the number of rapidly decreasing charge changes is increased by one;

[0170] It should be noted that the battery charge change rate threshold 1 is used to determine whether the charge change is slow enough. That is, if the charge drop rate in a certain time period is lower than the battery charge change rate threshold 1, the current energy consumption is considered to be basically stable. The battery charge change rate threshold 2 is used to define the critical boundary between "slow drop" and "rapid drop". If the charge drop rate in a certain period is higher than the battery charge change rate threshold 2, it is considered that the current energy consumption is too fast, and there may be high load or system abnormality. The battery charge change rate threshold 1 and the battery charge change rate threshold 2 are set by those skilled in the art based on the upper limit of the safe discharge rate supported by the battery type and environmental factors.

[0171] For example, under standard speed and medium mission load conditions, the average battery degradation rate of a certain type of unmanned vessel is 0.3% / min; under strong interference, high waves, and high mission load conditions, the average battery degradation rate is 1.2% / min. The battery charge change rate threshold 1 can be set to 0.5% / min to identify a stable state of charge, and the battery charge change rate threshold 2 can be set to 1.0% / min to determine whether the battery is entering a rapidly declining state.

[0172] S403: Let t = t + 1. If t is less than or equal to G - 1, continue executing S401 to S402; if t is greater than G - 1, execute S404.

[0173] S404: Compare the sizes of the number of stable changes in power, the number of slowly decreasing changes in power, and the number of quickly decreasing changes in power. If the number of stable changes in power is the largest, the power change trend type is the stable change type; if the number of slowly decreasing changes in power is the largest, the power change trend type is the slowly decreasing change type; if the number of quickly decreasing changes in power is the largest, the power change trend type is the quickly decreasing change type.

[0174] The method for calculating the battery charge change rate includes:

[0175]

[0176] Among them, DLBH t，t+1 is the battery charge change rate between the tth and t+1th time points, DCDL t+1 is the battery power at time point t+1, DCDL t is the battery power at the t-th time point, and Δt is the time interval between the t-th and t+1-th time points.

[0177] It should be noted that by continuously analyzing the battery charge change rate of N unmanned vessels over multiple time periods and combining it with a set change rate classification threshold, a mechanism for determining the type of battery charge change trend was established. This effectively identifies the battery consumption trend of the current unmanned vessel during mission execution. Complex continuous energy consumption behaviors were abstracted into three types: "stable," "slowly decreasing," or "rapidly decreasing." This provides efficient, clear, and predictive input for subsequent decisions such as mission scheduling, return determination, and energy compensation path optimization for the unmanned vessels.

[0178] The calculation method of the navigation stability index of the nth unmanned ship includes:

[0179] DHWD n =λ1×SBZC n +λ2×XBZC n ;

[0180]

[0181] Among them, DHWD n is the navigation stability index of the nth unmanned ship, SBZC n is the variance of the navigation speed of the nth unmanned ship, XBZC n is the heading variance of the nth unmanned ship, λ1 is the weighted coefficient of the navigation speed variance, λ2 is the weighted coefficient of the heading variance, HXSD g is the sailing speed at the g-th time point, HX g is the heading at the gth time point. λ1 and λ2 reflect the importance of the navigation speed variance and heading variance to the navigation stability index. For example, λ1 can be set to 0.6 and λ2 to 0.4. The navigation stability index is used to assess whether the UAV's heading is stable and whether there is any significant operational deviation due to sea conditions or wind interference.

[0182] The scheduling decision module is used to input the environmental perception feature data and state feature data corresponding to N unmanned ships into the scheduling decision level setting model to obtain the corresponding scheduling decision level. The scheduling decision levels include level one scheduling level, level two scheduling level, and level three scheduling level. Among them, level one scheduling level indicates that the unmanned ship is currently in an unsustainable observation state, with serious energy consumption risks, regional deviations, or stability anomalies, and requires immediate scheduling response operations; level two scheduling level indicates that the unmanned ship is currently in an operational but unstable state, with certain risk trends or local anomalies, and requires minor scheduling; level three scheduling level indicates that the unmanned ship is currently in a stable operating state, with good environmental and status conditions, and is suitable for continuing to perform the current observation task.

[0183] The training method of the scheduling decision level setting model includes:

[0184] Pre-collect a scheduling decision level setting data set, wherein the scheduling decision level setting data set includes DJ group scheduling decision level setting data and a scheduling decision level corresponding to the DJ group scheduling decision level setting data, where DJ is a positive integer greater than 0, and the scheduling decision level setting data includes environmental perception feature data and state feature data; divide the scheduling decision level setting data set into a training set and a validation set, wherein the training set is used to train a scheduling decision level setting model, and the validation set is used to evaluate the generalization performance of the scheduling decision level setting model;

[0185] During the training process of the scheduling decision level setting model, minimizing the cross-entropy loss function is used as the optimization objective. An early stopping strategy is used to monitor the performance of the validation set, and the network parameters are continuously adjusted to optimize the model performance. When the prediction accuracy on the validation set reaches the expected accuracy, the scheduling decision level setting model is considered to have converged and training is stopped. The scheduling decision level setting model is trained using a deep neural network based on a multilayer perceptron.

[0186] The scheduling decision level setting data is converted into feature vectors; the input layer of the scheduling decision level setting model receives the feature vectors, and the nonlinear relationship in the data is extracted through the hidden layer. Finally, the output layer of the scheduling decision level setting model calculates the probability distribution of the scheduling decision level through the softmax activation function, and outputs the scheduling decision level corresponding to the maximum probability as the final prediction result.

[0187] It should be noted that, in one embodiment of the present invention, in order to realize intelligent judgment of the dispatching level of multiple unmanned ships under complex environmental conditions, the system constructs a dispatching decision level setting model based on environmental perception feature data and state feature data. The model is trained by supervised learning, and the input end includes a multi-dimensional feature vector, and the output end is a corresponding dispatching decision level label, which is used to classify the dispatching control level that the unmanned ship is currently in. When training the dispatching decision level setting model, the environmental perception feature data includes the mean ocean current speed, the ocean current speed stability index, the mean ocean current direction, the ocean current direction stability index, the solar energy availability index and the wind energy disturbance level; the state feature data includes the unmanned ship deviation mark, the power change trend type and the navigation stability index.

[0188] The mean ocean current speed reflects the level of resistance during propulsion; the ocean current speed stability index is used to measure the amplitude of current speed fluctuation; the mean ocean current direction is used to evaluate the need for heading correction; the ocean current direction stability index is used to reflect the intensity of directional disturbance; the solar energy availability index is used to reflect the future endurance of the unmanned ship; the wind energy disturbance level is used to evaluate the degree of interference of the wind field on the attitude and track; the unmanned ship deviation mark is used to indicate whether there is a trend of crossing the boundary or approaching the boundary; the power change trend type is used to indicate the rate of change of battery power in a continuous time period (such as stable, slowly decreasing, and rapidly decreasing); the navigation stability index is used to measure the stability of the heading, and the higher the value, the more unstable the attitude.

[0189] Both environmental perception feature data and state feature data are normalized to the numerical range [0, 1] or mapped to classification vectors to construct a standard input feature vector. The system inputs the feature vectors of multiple samples and their corresponding labels (dispatch levels: first-level dispatch level, second-level dispatch level, and third-level dispatch level) into the dispatch decision level setting model for training. After the dispatch decision level setting model is trained, the operation phase only requires real-time extraction of the current unmanned vessel's environmental and state features and input into the model to output the dispatch decision level corresponding to the current moment. The dispatch decision level is used to guide whether to adjust the area, adjust parameters, or relay tasks for the current unmanned vessel, ensuring that the dispatch control has state responsiveness, environmental adaptability, and classification operability.

[0190] It should be further explained that the scheduling decision level setting model is based on preset multi-dimensional evaluation criteria and hierarchical judgment logic, and jointly analyzes the key characteristics of each unmanned ship, such as environmental adaptability, energy status, navigation stability and spatial position offset, to output the scheduling decision level corresponding to each unmanned ship. The scheduling decision level is used to comprehensively evaluate the current mission adaptability and operation stability of each unmanned ship.

[0191] In this application, the scheduling decision level setting model serves as the core intelligent judgment component. Based on the environmental perception feature data and the state feature data of the unmanned ship in the observation area, it dynamically determines whether each unmanned ship is currently suitable to continue to perform the observation task, and outputs the corresponding scheduling level (first-level scheduling, second-level scheduling, or third-level scheduling) accordingly. The introduction of this model breaks through the traditional scheduling method based on static rules or fixed threshold settings, and can achieve adaptive recognition and intelligent classification of the operating status of unmanned ships in complex and dynamic marine environments.

[0192] Specifically, through quantitative analysis of environmental perception feature data and state feature data, the scheduling decision level setting model can comprehensively evaluate whether a single ship has operational risks such as rapid decline in energy consumption, unstable navigation or observation deviation, thereby accurately identifying the scheduling level that requires immediate replacement (first-level scheduling), requires slight adjustment (second-level scheduling) or is suitable for maintaining operation (third-level scheduling).

[0193] The deployment of the scheduling decision level setting model not only improves the timeliness and accuracy of the system's identification of abnormal ships, but also provides a clear and structured input basis for subsequent successor scheduling or parameter adjustment strategies, realizing the transformation of unmanned ship cluster scheduling from "static preset" to "dynamic closed loop", and significantly enhancing the system's stable operation capability and resource coordination efficiency in complex and volatile sea environments.

[0194] The intelligent scheduling module intelligently schedules N unmanned ships based on their corresponding environmental perception feature data, state feature data and scheduling decision levels.

[0195] like Figure 4 As shown in FIG, the method for intelligently dispatching N unmanned ships includes:

[0196] The unmanned ships corresponding to the first-level scheduling decision level are constructed into a first-level scheduling unmanned ship set;

[0197] The unmanned ships corresponding to the second-level scheduling decision level are constructed into a second-level scheduling unmanned ship set, and the unmanned ships in the second-level scheduling unmanned ship set are intelligently scheduled;

[0198] The unmanned ships corresponding to the third-level scheduling decision level are constructed into a three-level scheduling unmanned ship set;

[0199] Based on the first-level scheduling unmanned ship set and the third-level scheduling unmanned ship set, the corresponding environmental perception feature data and state feature data are analyzed and processed to obtain a set of replacement candidate triples;

[0200] The unmanned ships in the first-level scheduling unmanned ship set and the second-level scheduling unmanned ship set are intelligently scheduled based on the replacement candidate triplet set.

[0201] The method for obtaining the replacement candidate triple set includes:

[0202] S500: The number of unmanned ships in the first-level scheduling unmanned ship set is recorded as YJ, and the number of unmanned ships in the third-level scheduling unmanned ship set is recorded as SJ; let the initial value of yj be 1, and the value range of yj is 1 to YJ; let the initial value of sj be 1, and the value range of sj is 1 to SJ;

[0203] S501: constructing the environmental perception feature data and state feature data of the yj-th unmanned ship in the first-level scheduling unmanned ship set into the first replacement fitness data; constructing the environmental perception feature data and state feature data of the sj-th unmanned ship in the third-level scheduling unmanned ship set into the second replacement fitness data;

[0204] S502: Calculate the succession fitness between the yj-th unmanned ship and the sj-th unmanned ship based on the first succession fitness data and the second succession fitness data; construct a succession candidate triplet by using yj, sj and the succession fitness data, and add the succession candidate triplet to the succession candidate triplet set;

[0205] S503: Let sj = sj + 1. If sj is less than or equal to SJ, continue executing S501 to S502. If sj is greater than SJ, let yj = yj + 1. If yj is less than or equal to YJ, let sj = 1 and continue executing S501 to S502. If yj is greater than YJ, obtain the successor candidate triple set and end the current process.

[0206] The method for obtaining the replacement suitability between the yj-th unmanned ship and the sj-th unmanned ship includes:

[0207]

[0208] Among them, SPD yj，sj is the succession fitness between the yj-th unmanned ship and the sj-th unmanned ship, ZXJL(yj, sj) represents the distance between the yj-th unmanned ship and the sj-th unmanned ship, TYHQ sj Represents the solar energy availability index of the sjth unmanned ship, WDI sj Represents the wind energy disturbance level index of the sjth unmanned ship, DHWD sj represents the navigation stability index of the sjth unmanned ship. μ1, μ2, μ3, and μ4 are weighting coefficients, μ1 + μ2 + μ3 + μ4 = 1. For example, μ1 can be set to 0.25, μ2 to 0.25, μ3 to 0.25, and μ4 to 0.25.

[0209] is the distance factor (positive correlation), the closer the distance between the yj-th unmanned ship and the sj-th unmanned ship, the shorter the task migration path and the lower the scheduling response time. sj is the solar energy factor (positive correlation), TYHQ sj Reflects the energy supply capacity of the sjth unmanned ship during the relay mission. The better the lighting conditions and the stronger the power supply capacity, the longer the sjth unmanned ship can support the mission. Therefore, the larger the solar factor, the higher the adaptability. μ3×WDI sj is the wind field interference factor (negative correlation). The larger the wind field interference factor, the stronger the wind disturbance, the lower the path tracking and heading stability, and the worse the adaptability. Therefore, the wind field interference factor is a negative coefficient. sj Navigation stability factor (negative correlation): DHWD sj The larger the value, the more unstable the navigation, the higher the mission risk, and the lower the adaptability. Therefore, the navigation stability factor is a negative coefficient. By combining the distance factor, solar energy factor, wind field interference factor, and navigation stability factor to form a multi-dimensional evaluation of candidate unmanned vessels, the mission migration strategy can be more closely aligned with the comprehensive operational capabilities of the unmanned vessel and the current environmental conditions.

[0210] like Figure 5 As shown, the method for intelligently scheduling the unmanned ships in the first-level scheduling unmanned ship set and the second-level scheduling unmanned ship set based on the replacement candidate triple set includes:

[0211] S600: Obtaining the succession candidate triplet with the highest succession fitness from the succession candidate triplet set, and recording it as the current succession candidate triplet; the first digit of the succession candidate triplet is the serial number of the unmanned ship to be replaced; the second digit of the succession candidate triplet is the serial number of the candidate succession unmanned ship; the third digit of the succession candidate triplet is the succession fitness when the unmanned ship corresponding to the second digit replaces the unmanned ship corresponding to the first digit;

[0212] S601: Obtain the observation area of ​​the unmanned ship corresponding to the first position of the candidate triplet, which is recorded as the first observation area; obtain the observation area of ​​the unmanned ship corresponding to the second position of the candidate triplet, which is recorded as the second observation area;

[0213] S602: Multiply the area of ​​the first observation area by a preset observation area replacement ratio (for example, the observation area replacement ratio can be set to one-fifth) to obtain an apportioned observation area area; subtract the first observation area from the apportioned observation area to obtain a first updated observation area area, where the first updated observation area area is the updated observation area area of ​​the unmanned ship corresponding to the first position of the candidate triplet; sum the second observation area and the apportioned observation area area to obtain a second updated observation area area, where the second updated observation area area is the updated observation area area of ​​the unmanned ship corresponding to the second position of the candidate triplet;

[0214] Optionally, the system can dynamically adjust the takeover ratio based on the power level, stability and other status characteristics of the relay unmanned ship to improve the sustainability and load balance of the task relay.

[0215] S603: Remove all successor candidate triples corresponding to the first or second position in the current successor candidate triple from the successor candidate triple set to prevent a single unmanned ship from repeatedly participating in the scheduling relay;

[0216] S604: Repeat S600 to S603, and stop executing when the set of successor candidate triples is empty.

[0217] The method for intelligently scheduling the unmanned ships in the secondary scheduling unmanned ship set includes:

[0218] S700: Record the number of unmanned ships in the secondary scheduling unmanned ship set as EJ; let the initial value of ej be 1, and the value range of ej be 1 to EJ;

[0219] S701: Obtain the ejth unmanned ship from the secondary scheduling unmanned ship set, input the environmental perception feature data and state feature data of the ejth unmanned ship into the unmanned ship parameter setting model, and obtain the unmanned ship adjustment parameters corresponding to the ejth unmanned ship, which are recorded as the unmanned ship target adjustment parameters; the unmanned ship adjustment parameters include propulsion power, rudder angle adjustment frequency, and attitude control accuracy;

[0220] S702: Adjust the current state parameters of the ejth unmanned ship to the target adjustment parameters of the unmanned ship; thereby promoting the unmanned ship from the second-level scheduling level to the third-level scheduling level, that is, from an operational but unstable state to a stable state;

[0221] S703: Let ej=ej+1. If ej is less than or equal to EJ, continue executing S701 to S702; if ej is greater than EJ, end the current process.

[0222] The training method of the unmanned ship parameter setting model includes:

[0223] Pre-constructing an unmanned ship parameter setting data set, the unmanned ship parameter setting data set including a CS group of unmanned ship parameter setting data and unmanned ship adjustment parameters corresponding to the CS group of unmanned ship parameter setting data, where CS is a positive integer greater than 0, and the unmanned ship parameter setting data includes environmental perception feature data and state feature data; dividing the unmanned ship parameter setting data set into an unmanned ship parameter setting data training set and an unmanned ship parameter setting data validation set, wherein the unmanned ship parameter setting data training set is used for parameter learning of the unmanned ship parameter setting model, and the unmanned ship parameter setting data validation set is used for real-time evaluation of the generalization ability of the unmanned ship parameter setting model;

[0224] During the training process of the unmanned ship parameter setting model, a deep neural network structure based on multi-layer perceptron is adopted to convert the unmanned ship parameter setting data into a feature vector as input, extract the nonlinear features in the data through the hidden layer, and finally use the softmax activation function in the output layer to generate the probability distribution of the unmanned ship adjustment parameters, and output the unmanned ship adjustment parameters corresponding to the maximum probability as the final prediction result; the training process aims to minimize the cross-entropy loss function, and at the same time introduces an early stopping strategy to monitor the performance of the unmanned ship parameter setting data verification set. When the prediction accuracy on the unmanned ship parameter setting data verification set reaches the preset accuracy, the unmanned ship parameter setting model is considered to have converged and the training stops immediately.

[0225] It should be noted that propulsion power controls the speed of the unmanned vessel and its energy consumption per unit time. High speeds and high-frequency starts will cause the battery power to decrease faster, putting the unmanned vessel in an energy-risk state. The frequency of rudder angle adjustment directly affects the navigation stability index, that is, the directional fluctuation during navigation. Frequent and excessive rudder angle changes will cause the hull to swing and the route to oscillate, especially in windy and wavey areas, which will be judged as "unstable navigation." In areas of wind and wave interference, the hull of the unmanned vessel will tilt or shake slightly, affecting the stability of the instrument and navigation of the unmanned vessel. If the sampling period corresponding to the attitude control accuracy is too long or the correction frequency is too low, it will be judged as "unstable operation."

[0226] Example 2

[0227] See also Figure 3 As shown, this embodiment provides an unmanned ship real-time scheduling and control system, which also includes:

[0228] The ratio setting module inputs the environmental perception feature data and state feature data of the unmanned ship to be replaced into the replacement ratio setting model to obtain a dynamically set observation area replacement ratio.

[0229] The training method of the succession ratio setting model includes:

[0230] Pre-constructing a succession ratio setting data set, the succession ratio setting data set including X groups of succession ratio setting data and succession ratios of observation areas corresponding to the X groups of succession ratio setting data, where X is a positive integer greater than 0, and the succession ratio setting data including environmental perception feature data and state feature data; dividing the succession ratio setting data set into a succession ratio setting data training set and a succession ratio setting data validation set, wherein the succession ratio setting data training set is used for parameter learning of a succession ratio setting model, and the succession ratio setting data validation set is used for real-time evaluation of the generalization ability of the succession ratio setting model;

[0231] During the training process of the succession ratio setting model, a deep neural network structure based on multi-layer perceptron is adopted to convert the succession ratio setting data into a feature vector as input, extract the nonlinear features in the data through the hidden layer, and finally use the softmax activation function in the output layer to generate the probability distribution of the succession ratio of the observation area, and output the succession ratio of the observation area corresponding to the maximum probability as the final prediction result; the training process aims to minimize the cross-entropy loss function, and at the same time introduces an early stopping strategy to monitor the performance of the succession ratio setting data validation set. When the prediction accuracy on the succession ratio setting data validation set reaches the preset accuracy, the succession ratio setting model is considered to have converged and the training stops immediately.

[0232] Example 3

[0233] See also Figure 2 As shown, this embodiment provides a real-time scheduling control method for an unmanned ship, including:

[0234] Based on the environmental perception data within the target observation sea area and the state parameters of N unmanned ships within a unit time, the observation areas corresponding to the N unmanned ships are pre-constructed;

[0235] Based on the observation areas corresponding to the N unmanned ships, feature extraction is performed on the environmental perception data within the target observation sea area per unit time, and the environmental perception feature data corresponding to the N unmanned ships are obtained;

[0236] Feature extraction is performed based on the state parameters of N unmanned ships per unit time, and state feature data corresponding to the N unmanned ships is obtained;

[0237] The environmental perception feature data and state feature data corresponding to N unmanned ships are respectively input into the scheduling decision level setting model to obtain the corresponding scheduling decision level;

[0238] N unmanned ships are intelligently dispatched based on their corresponding environmental perception feature data, state feature data and dispatch decision levels.

[0239] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any modifications or substitutions that can be easily conceived by a person skilled in the art within the technical scope disclosed in the present invention should be included in the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be based on the scope of protection of the claims.

[0240] Finally: The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A real-time dispatching and control system for unmanned ships, characterized in that: include: The area construction module pre-constructs the observation area corresponding to N unmanned ships based on the environmental perception data in the target observation sea area per unit time and the status parameters of N unmanned ships; The first processing module extracts features of the environmental perception data within the target observation sea area per unit time based on the observation areas corresponding to the N unmanned ships, and obtains the environmental perception feature data corresponding to the N unmanned ships; The second processing module extracts features based on the state parameters of the N unmanned ships per unit time to obtain state feature data corresponding to the N unmanned ships; The scheduling decision module is used to input the environmental perception feature data and state feature data corresponding to N unmanned ships into the scheduling decision level setting model to obtain the corresponding scheduling decision level; The intelligent scheduling module intelligently schedules N unmanned ships based on their corresponding environmental perception feature data, state feature data and scheduling decision levels.

2. The unmanned ship real-time dispatching and control system according to claim 1, characterized in that: The scheduling decision level includes a first-level scheduling level, a second-level scheduling level and a third-level scheduling level; The method for intelligently dispatching N unmanned ships includes: The unmanned ships corresponding to the first-level scheduling decision level are constructed into a first-level scheduling unmanned ship set; The unmanned ships corresponding to the second-level scheduling decision level are constructed into a second-level scheduling unmanned ship set, and the unmanned ships in the second-level scheduling unmanned ship set are intelligently scheduled; The unmanned ships corresponding to the third-level scheduling decision level are constructed into a three-level scheduling unmanned ship set; Based on the first-level scheduling unmanned ship set and the third-level scheduling unmanned ship set, the corresponding environmental perception feature data and state feature data are analyzed and processed to obtain a set of replacement candidate triples; The unmanned ships in the first-level scheduling unmanned ship set and the second-level scheduling unmanned ship set are intelligently scheduled based on the replacement candidate triplet set.

3. The unmanned ship real-time dispatching and control system according to claim 2, characterized in that: The method for intelligently scheduling the unmanned ships in the secondary scheduling unmanned ship set includes: Obtain each unmanned ship in the secondary scheduling unmanned ship set, input the corresponding environmental perception feature data and state feature data into the unmanned ship parameter setting model, and obtain the corresponding unmanned ship target adjustment parameters. The unmanned ship target adjustment parameters include propulsion power, rudder angle adjustment frequency and attitude control accuracy; based on the unmanned ship target adjustment parameters, adjust the current state parameters of the unmanned ship to upgrade the unmanned ship from the secondary scheduling level to the tertiary scheduling level.

4. The unmanned ship real-time dispatching and control system according to claim 2, characterized in that: The method for obtaining the replacement candidate triple set includes: The environmental perception feature data and state feature data corresponding to the first-level dispatched unmanned ship and the third-level dispatched unmanned ship are constructed as the replacement fitness data, and the replacement fitness between each group of first-level and third-level unmanned ships is calculated based on the data; According to the succession fitness result, a triple containing the serial number of the unmanned ship to be replaced, the serial number of the candidate succession unmanned ship and the corresponding fitness is constructed to form a succession candidate triple set.

5. The unmanned ship real-time dispatching and control system according to claim 4, characterized in that: The method for intelligently scheduling unmanned ships in the first-level scheduling unmanned ship set and the second-level scheduling unmanned ship set based on the replacement candidate triple set includes: Obtain the candidate triplet with the highest succession fitness. Based on the observation area of ​​the unmanned ship to be replaced and the candidate unmanned ship contained in the candidate triplet, recalculate the updated observation area of ​​the two according to the preset observation area succession ratio, and use them as the scheduling optimization results. All candidates involved in the used successor candidate triples are removed, and the above operation is repeated for the remaining successor candidate triples until the successor candidate triple set is empty.

6. The unmanned vessel real-time dispatching and control system according to claim 1, characterized in that: The method of pre-constructing the observation area corresponding to N unmanned ships includes: Acquire environmental perception data within the target observation sea area, the environmental perception data including ocean current velocity field, ocean current direction map, wind speed, wind direction, wave level and solar radiation intensity; Obtaining status parameters of N unmanned ships, including real-time position coordinates, battery power, navigation speed, and heading; The state parameters and environmental perception data corresponding to N unmanned ships are input into the comprehensive feasibility evaluation model respectively to obtain the comprehensive feasibility scores corresponding to the N unmanned ships; The comprehensive feasibility scores corresponding to the N unmanned ships are normalized to obtain the regional allocation factors corresponding to the N unmanned ships. Combined with the total area of ​​the target observation sea area, the observation area that the N unmanned ships should undertake is calculated; the observation area that the N unmanned ships should undertake is the observation area corresponding to the N unmanned ships.

7. The unmanned vessel real-time dispatching and control system according to claim 1, characterized in that: The method for obtaining the environmental perception feature data corresponding to N unmanned ships includes: Obtain the environmental perception data of the observation area corresponding to each unmanned vessel, calculate the mean current velocity and the current velocity stability index of the current velocity field, and calculate the mean current direction and the current direction stability index of the current direction map; Calculate the solar energy availability index based on the solar radiation intensity, obtain the wind speed disturbance degree and wind direction jump frequency index based on the wind speed and wind direction data, and obtain the wind energy disturbance level based on the preset wind energy disturbance level mapping table; The indicators calculated above are constructed into the environmental perception feature data of each unmanned ship.

8. The unmanned vessel real-time dispatching and control system according to claim 1, characterized in that: The method for obtaining the state characteristic data corresponding to N unmanned ships includes: Obtain the real-time status parameters of each unmanned vessel, identify the deviation mark of the unmanned vessel by analyzing the change trend of the position coordinates, analyze the change trend of the battery power to obtain the type of power change trend, and calculate the navigation stability index based on the navigation speed and heading stability; The unmanned ship deviation mark, power change trend type and navigation stability index are constructed into the status characteristic data corresponding to each unmanned ship.

9. The unmanned ship real-time dispatching and control system according to claim 8, characterized in that: The method for obtaining the unmanned ship deviation mark of the nth unmanned ship includes: The closest distance between the historical trajectory point of the unmanned ship and the boundary of its observation area is calculated, and the proportion of out-of-bounds trajectory points is counted; if the proportion exceeds the preset out-of-bounds ratio threshold, the unmanned ship deviation flag is set to yes, otherwise it is set to no.

10. The unmanned ship real-time dispatching and control system according to claim 9, characterized in that: The method for obtaining the power change trend type of the nth unmanned ship includes: Calculate the battery charge change rate based on the battery charge at consecutive time points. Combined with the preset battery charge change rate threshold 1 and battery charge change rate threshold 2, count the number of stable charge changes, slow charge decreases, and rapid charge decreases respectively. The change type with the largest number of changes is taken as the power change trend type of the nth unmanned ship.

Images