Self-organized electric quantity supply method and system for unmanned ship

By setting communication channels and offshore battery swap platforms in unmanned fleets, and using the power optimization model to schedule inspection and power replenishment tasks of unmanned ships, the problem of high cost and frequent charging of unmanned fleets in remote sea environments is solved, and the problem of high cost and frequent charging of unmanned fleets is improved. The inspection and monitoring efficiency and continuous operation ability are achieved.

CN118917646AActive Publication Date: 2024-11-08TIMES TIANHAI (XIAMEN) INTELLIGENT TECH CO LTD

Patent Information

Application Number
CN202411416387.9
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-10-11
Publication Date
2024-11-08
Estimated Expiration
2044-10-11

AI Technical Summary

Technical Problem

In the existing unmanned fleets in patrol and surveillance applications in remote sea environments, frequent charging is caused by relying on high-cost satellite communications and the difference in power consumption between ships, which affects patrol efficiency.

Method used

By setting communication channels and offshore battery swap platforms, self-organized unmanned fleets to optimize power supply and data transmission, and use the power optimization model to schedule unmanned ship inspection and power supply tasks to achieve dynamic power coordination and data transmission efficiency improvement.

Benefits of technology

It effectively reduces the operating costs of unmanned fleets, improves patrol and monitoring efficiency and continuous operation capabilities, and is suitable for tasks in long-term and complex environments in the sea.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118917646A_ABST
    Figure CN118917646A_ABST
Patent Text Reader

Abstract

The invention relates to the technical field of self-organized power supply of unmanned intelligent fleet, and discloses a self-organized power supply method and system for unmanned ships, and the method comprises the steps: setting a communication channel between a power supply port and a cruise area, and determining the number of unmanned ships A according to the set communication channel; the operation area is divided into a plurality of sub-areas, and the number of unmanned ships B and unmanned ships C in the operation area is determined according to the number of the sub-areas; collecting state information of the unmanned ship B and the unmanned ship C, and constructing an electric quantity optimization model; on the basis of the electric quantity optimization model, patrol monitoring and power supply scheduling are carried out on the unmanned ship C; the method can effectively improve the cruising ability of the unmanned ship and reduce the cost of data transmission, and is suitable for long-time open-sea operation.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the field of self-organized power supply of unmanned intelligent fleets, and more specifically, to a method and system for self-organized power supply of unmanned ships. Background Art

[0002] With the continuous advancement of technology, unmanned fleets are widely used in ocean patrol and monitoring.

[0003] However, the existing unmanned fleet still has the following problems in practical application:

[0004] At present, some unmanned fleets rely on satellite communications for data transmission. Although satellite communications can cover a wide range of sea areas, their charges are relatively high, which significantly increases the operating costs of unmanned fleets and limits the application of unmanned fleets in patrol and monitoring in offshore environments. In addition, since unmanned ships perform missions in different sea areas, there are large differences in power consumption between ships, causing some ships to frequently return to ports for charging, affecting patrol and monitoring efficiency. Summary of the invention

[0005] The present invention provides a method and system for self-organizing power supply of an unmanned ship, which solves the technical problems raised in the background technology.

[0006] In a first aspect, a method for self-organizing power supply of an unmanned ship comprises:

[0007] Step S101, setting a communication channel between the power supply port and the cruising area, and determining the number of unmanned ships A according to the set communication channel; the unmanned ship A is used to obtain fully charged batteries from the power supply port and transport them to the offshore power exchange platform, and maintain the data connection between the offshore power exchange platform and the power supply port;

[0008] Among them, the offshore power exchange platform and the power supply port are respectively used as the two ends of the communication channel. The offshore power exchange platform is an offshore floating platform at an anchored position. Both the offshore power exchange platform and the power supply port are equipped with equipment for automatic power exchange for unmanned ships.

[0009] Step S102, dividing the operation area into several sub-areas, and determining the number of unmanned ships B and unmanned ships C in the operation area according to the number of sub-areas; unmanned ship B is used to receive fully charged batteries from the offshore power exchange platform and maintain data connection between unmanned ship C and the offshore power exchange platform; unmanned ship C is used to perform patrol monitoring tasks;

[0010] Step S103, collecting status information of the unmanned boat B and the unmanned boat C, and constructing a power optimization model; wherein the status information includes: power consumption, driving speed, driving direction, wind force, wind force direction, water flow force and water flow direction data;

[0011] Step S104: Based on the power optimization model, the unmanned boat C is patrolled and monitored and the power supply is scheduled.

[0012] Furthermore, according to the set communication channel, the number of unmanned ships A is determined, including:

[0013] Collect the length of the communication channel And battery replacement operation time ; Among them, the length of the communication channel Indicates the length of the path set between the power supply port and the offshore power exchange platform;

[0014] Determine the maximum moving speed of unmanned ship A is , moving speed represents the maximum value of the intersection of the speed ranges of all unmanned ships A;

[0015] pass Determine the round trip time of unmanned ship A in the communication channel ; Based on round trip time Determine the unmanned ship A at any time during the operation period The coordinate position is calculated as follows:

[0016] When the unmanned ship A moves, the coordinates ;

[0017] When the unmanned ship A is stationary, the coordinates ;in, and They represent the time when the unmanned ship A arrives at the power supply port and the offshore power exchange platform, respectively. It indicates the operation time of the unmanned ship A after it joins the communication channel from the power supply port. It represents the operation of converting the distance traveled by the unmanned ship A in the communication channel into the coordinates of the unmanned ship A based on the coordinates of the communication channel;

[0018] pass Determine the signal coverage of unmanned ship A ; Wherein, h represents the maximum antenna height of the unmanned ship A; based on the coordinates of the unmanned ship A at any time and the wireless signal coverage range, the coverage range of the unmanned ship A at any time during the operation period is obtained; the number of unmanned ships A is gradually increased iteratively, and the number of unmanned ships A is determined to be N;

[0019] In the operation time period, a time period is At any time, determine the union of the wireless signal coverage intervals of N unmanned ships A to cover the communication channel.

[0020] Furthermore, the operation area is divided into a number of sub-areas, and the number of unmanned ships B and unmanned ships C in the operation area is determined according to the number of sub-areas, including:

[0021] Divide the operation area into several sub-areas and assign an unmanned ship C to each sub-area;

[0022] The maximum distance between each sub-region and its adjacent sub-region is obtained by brute force method, and the number of maximum distances is obtained according to the number of adjacent sub-regions;

[0023] Based on the ultrashort wave propagation distance combined with K-means clustering of the maximum distance, several signal missing areas are obtained according to the distribution of the maximum distance. The signal missing area is a polygon constructed by the maximum distance.

[0024] The geometric center of the signal missing area is determined by the centroid algorithm, and the unmanned ship B is gradually and iteratively filled around the geometric center of the signal missing area to ensure that the signal range of the unmanned ship B covers the signal missing area.

[0025] Furthermore, after the unmanned ship C collects data from the sea area of ​​the operation area, it transmits the data to the power supply port through wireless signals, including:

[0026] Construct a tree graph, map the offshore power exchange platform as the root node, map the unmanned ship B and the unmanned ship C as child nodes, and classify the unmanned ship B and the unmanned ship C according to the depth of the node to obtain a tree graph;

[0027] The tree diagram is updated at fixed time intervals, and the unmanned ships B and C dynamically adjust the node classification according to the depth of the corresponding nodes; the depth of the node is determined by the minimum number of hops that the node data passes through to reach the root node;

[0028] Wherein, in the tree diagram, multiple n-level nodes and multiple n-1-level nodes all establish data connection relationships, and the multiple n-level nodes are evenly distributed to multiple n-1-level nodes;

[0029] Transmit data in slices according to the tree graph and time window size at any moment;

[0030] pass Determine the amount of data transmitted by each secondary node at any time; where, Indicates the data transmission capacity of the wireless signal at any time. The number of secondary nodes connected to the root node determines the amount of data transmitted by each secondary node at any time; Indicates the data transmission capacity of the wireless signal at any time. Indicates the number of secondary nodes connected to the root node;

[0031] pass Determine the data transmission volume of each third-level node at any time; where, Indicates the number of child nodes of any secondary node, Indicates the type of the secondary node; if the type of the secondary node is unmanned ship C, then is 1, and unmanned ship B is is 0; by looping this process, the data transmission size per unit time of each node is obtained;

[0032] At any time, if the i-th node in the tree diagram is the unmanned ship C, then determine whether the i-th node is a leaf node; if the i-th node is a leaf node, then the data slice size of the i-th node is equal to the data transmission volume of the i-th node at any time; if the i-th node is an intermediate node, then the data slice size of the i-th node is equal to the data transmission volume of the child node of the i-th node at any time;

[0033] The slice data of each time window are marked with timestamps. Unmanned ships A, B and C use timestamps as priority when transmitting, and give priority to transmitting slice data with earlier timestamps; if the timestamps of unmanned ships A, B and C are the same, the smaller slice data will be transmitted first.

[0034] Furthermore, the status information of the unmanned ship B and the unmanned ship C is collected to build a power optimization model, including:

[0035] The power consumption, driving speed, driving direction, wind force, wind direction, water flow and water flow direction data of the unmanned boat B and the unmanned boat C are collected at fixed time intervals t; the water flow and water flow direction are measured by the monitoring units installed on the hulls of the unmanned boats B and C;

[0036] The driving speed, driving direction, wind force, wind direction, water flow and water flow direction data are normalized and constructed into training samples, and the power consumption is normalized as the sample label of the training sample; thus, a training set of the power optimization model is obtained;

[0037] The power optimization model includes an input layer and several hidden layers. The input layer is used to input the feature vector constructed after normalizing the driving speed, water flow size and water flow direction data; the hidden layer is used to output the updated state of the feature vector, and the updated state is output to the classifier. The classification space of the classifier represents the power consumption based on the driving speed, driving direction, wind force, wind direction, water flow size and water flow direction in a time interval t; the power optimization model reversely updates the weight parameters and bias parameters of the hidden layer through the mean error variance loss function, and determines the convergence of the power optimization model through cross-validation.

[0038] Furthermore, based on the power optimization model, the patrol and power replenishment scheduling of the unmanned ship C is carried out, including:

[0039] Collect the local weight of the unmanned ship C. The local weight represents the number of unmanned ships B and unmanned ships C that are disconnected from the data connection in the local network by simulating shielding the ultrashort wave signal of the unmanned ship C at any time. The unmanned ship C with a local weight of 0 and a power higher than a preset high power threshold is taken as a candidate ship, and the area where the candidate ship is located is taken as the candidate area;

[0040] The remaining power of unmanned ship B, unmanned ship C and offshore power exchange platform, and the coordinates of unmanned ship B and unmanned ship C are collected at fixed time intervals t; unmanned ship B with power below a preset low power threshold and unmanned ship C with power below a preset medium power threshold are respectively used as the first power supply ship and the second power supply ship;

[0041] Determine whether there is an overlapping unmanned ship C between the second power-supply ship and the backup ship;

[0042] If there are overlapping unmanned boats C, the overlapping unmanned boats C will directly replace part of the first power-supply boat, and the remaining unmanned boats C in the candidate boats will directly replace part of the second power-supply boat, and the replaced second power-supply boat will again replace part of the first power-supply boat;

[0043] If there is no overlapping unmanned ship C, the unmanned ship C in the candidate ship will directly replace part of the second power-supply ship, and the replaced second power-supply ship will replace part of the first power-supply ship again;

[0044] The replaced first power supply ship goes to the offshore unmanned platform to perform the task of replacing fully charged batteries. The replaced second power supply ship goes to the location of the first power supply ship to perform the data connection task. After replacing the fully charged batteries, the first power supply ship goes to the alternative area to perform patrol and monitoring tasks.

[0045] Furthermore, based on the power optimization model, the unmanned ship C is dispatched for patrol monitoring, including:

[0046] Collect the driving speed, driving direction, wind force, wind direction, water flow and water flow direction data of the unmanned ship B and the unmanned ship C at the i-th moment, determine the power consumption of the unmanned ship B and the unmanned ship C in the next time interval t based on the driving speed, driving direction, wind force, wind direction, water flow and water flow direction data, and obtain the remaining power of the unmanned ship B and the unmanned ship C at the i+t-th moment;

[0047] According to the wind force, wind direction, current force and current direction of unmanned ship B at the i-th moment, the number of the first power-supplying ships at the i+t-th moment is determined through the power optimization model;

[0048] Compare the number of candidate ships with the number of first replenishment ships;

[0049] If the number of candidate ships is less than or equal to the number of first power-supply ships, the power consumption of the unmanned ship C in the next time interval t is adjusted by adjusting the speed of the unmanned ship C, and the number of second power-supply ships at the i+tth moment is determined to be equal to the number of candidate ships;

[0050] If the number of candidate ships is greater than the number of first power-supply ships, the power consumption of the unmanned ship C in the next time interval t is adjusted by adjusting the speed of the unmanned ship C, and the number of second power-supply ships at the i+tth moment is determined to be equal to the number of first power-supply ships;

[0051] Among them, the driving speed of the unmanned boat C at the i-th moment is determined by combining the optimization algorithm with the power optimization model.

[0052] Furthermore, the driving speed of the unmanned ship C at the i-th moment is determined by combining the optimization algorithm with the power optimization model, including:

[0053] Randomly generate an initial particle swarm with R individuals that meet the constraints, and initialize the current number of iterations to 1;

[0054] Encode the individual, and each element of the code represents the driving speed , Driving direction , Wind force , Wind direction , water flow size and water flow direction Data; coded as follows: ; ; ;

[0055] in, and They represent the wind force and water flow size that the unmanned ship B is subjected to, and They represent the weight coefficient of wind force and the weight coefficient of water flow respectively, and All were obtained by linear least squares fitting;

[0056] The constraints are that the driving speed is not 0, the driving direction, wind direction and water flow direction do not change, and the wind force and water flow size satisfy the linear relationship with the driving speed and driving direction;

[0057] The fitness function of each individual is obtained through the power optimization model. The fitness function value represents the power consumption in the next time interval t. The calculation formula is as follows: in represents the output of the power optimization model for individual codes, Indicates driving speed The weight coefficient of

[0058] Sort the R individuals according to their fitness values, retain a fixed number of individuals, and update the remaining individuals as parents;

[0059] Among them, only the driving speed of the parent individual is updated, and the wind force and water flow are updated according to the updated driving speed;

[0060] Determine whether the current number of iterations is equal to the maximum number of iterations; if the current number of iterations = the maximum number of iterations, determine the speed of the unmanned ship C at the i-th moment , otherwise continue iterating.

[0061] Furthermore, unmanned ship B and unmanned ship C determine their unmanned ship types based on the tasks to be performed, and respond to scheduling based on the types of unmanned ships.

[0062] In a second aspect, a self-organizing power supply system for an unmanned ship includes:

[0063] Local area network module, used to build communication connections between the power supply port, unmanned power exchange platform and operation area;

[0064] The data transmission module is used to segment and timestamp the data collected by the unmanned ship C and transmit them according to the priority;

[0065] The scheduling module is used to perform power replenishment scheduling for unmanned ship C and unmanned ship B, and adjust the power consumption of unmanned ship C through a preset power optimization model.

[0066] The beneficial effects of the present invention are: based on the power optimization model, the unmanned fleet can work intelligently and collaboratively at sea, and the unmanned ships can be dynamically dispatched to replenish power, thus avoiding the congestion of power supply; based on the construction of communication channels and tree diagrams to optimize data transmission, the smooth and efficient data transmission of the unmanned fleet when performing tasks is ensured. The system effectively improves the operating efficiency and continuous cruising capability of the unmanned fleet, and is suitable for long-term operations in complex environments in the open sea. BRIEF DESCRIPTION OF THE DRAWINGS

[0067] Figure 1 This is a flow chart of a self-organized power supply method for an unmanned ship of the present invention;

[0068] Figure 2 It is a dispatch diagram of the unmanned ship B and the unmanned ship C of the present invention. DETAILED DESCRIPTION

[0069] The subject matter described herein will now be discussed with reference to example embodiments. It should be understood that the discussion of these embodiments is only to enable those skilled in the art to better understand and implement the subject matter described herein, and the functions and arrangements of the elements discussed may be changed without departing from the scope of protection of the contents of this specification. Each example may omit, replace or add various processes or components as needed. In addition, the features described relative to some examples may also be combined in other examples.

[0070] like Figure 1-2 As shown, a method for self-organizing power supply of an unmanned ship includes:

[0071] Step S101, setting a communication channel between the power supply port and the cruising area, and determining the number of unmanned ships A according to the set communication channel; the unmanned ship A is used to obtain fully charged batteries from the power supply port and transport them to the offshore power exchange platform, and maintain the data connection between the offshore power exchange platform and the power supply port;

[0072] Among them, the offshore power exchange platform and the power supply port are respectively used as the two ends of the communication channel. The offshore power exchange platform is an offshore floating platform at an anchored position. Both the offshore power exchange platform and the power supply port are equipped with equipment for automatic power exchange for unmanned ships.

[0073] Step S102, dividing the operation area into several sub-areas, and determining the number of unmanned ships B and unmanned ships C in the operation area according to the number of sub-areas; unmanned ship B is used to receive fully charged batteries from the offshore power exchange platform and maintain data connection between unmanned ship C and the offshore power exchange platform; unmanned ship C is used to perform patrol monitoring tasks;

[0074] Step S103, collecting status information of the unmanned boat B and the unmanned boat C, and constructing a power optimization model; wherein the status information includes: power consumption, driving speed, driving direction, wind force, wind force direction, water flow force and water flow direction data;

[0075] Step S104: Based on the power optimization model, the unmanned boat C is patrolled and monitored and the power supply is scheduled.

[0076] In one embodiment of the present invention, determining the number of unmanned ships A according to the set communication channel includes:

[0077] Collect the length of the communication channel And battery replacement operation time ; Among them, the length of the communication channel Indicates the length of the path set between the power supply port and the offshore power exchange platform;

[0078] Determine the maximum moving speed of unmanned ship A is , moving speed represents the maximum value of the intersection of the speed ranges of all unmanned ships A;

[0079] pass Determine the round trip time of unmanned ship A in the communication channel ; Based on round trip time Determine the unmanned ship A at any time during the operation period The coordinate position is calculated as follows:

[0080] When the unmanned ship A moves, the coordinates ;

[0081] When the unmanned ship A is stationary, the coordinates ;in, and They represent the time when the unmanned ship A arrives at the power supply port and the offshore power exchange platform, respectively. It indicates the operation time of the unmanned ship A after it joins the communication channel from the power supply port. It represents the operation of converting the distance traveled by the unmanned ship A in the communication channel into the coordinates of the unmanned ship A based on the coordinates of the communication channel;

[0082] pass Determine the signal coverage of unmanned ship A ; Wherein, h represents the maximum antenna height of the unmanned ship A; based on the coordinates of the unmanned ship A at any time and the wireless signal coverage range, the coverage range of the unmanned ship A at any time during the operation period is obtained; the number of unmanned ships A is gradually increased iteratively, and the number of unmanned ships A is determined to be N;

[0083] In the operation time period, a time period is At any time, determine the union of the wireless signal coverage intervals of N unmanned ships A to cover the communication channel.

[0084] For example, the length of the communication channel is set to L, the ultra-short wave frequency of the unmanned ship is kept in the very high frequency band, and the speed of the unmanned ship A is equal and the speed is v. ;

[0085] By formula Get the ultra-short wave coverage of each unmanned ship; then, use the open source network simulation framework OMNeT++ to simulate an unmanned ship A entering the communication channel and running a round-trip time , test whether the power supply port and the offshore power exchange platform have a continuous data connection; if not, when the first unmanned ship leaves the port terminal, Simulate the second unmanned ship A to enter the communication channel, so as to gradually increase the number of simulated unmanned ships A;

[0086] When the number of unmanned ships A is determined to be N, the power supply port and the offshore power exchange platform are exactly in a time period of If the connection is maintained continuously, it means that the number of unmanned ships A is the least and the communication channel can be built to connect the power supply port and the offshore power exchange platform. The following formula is used to simulate the position of unmanned ship A in the channel:

[0087] When the unmanned ship A moves, the coordinates ;

[0088] When the unmanned ship A is stationary, the coordinates ;in, and They represent the time when the unmanned ship A arrives at the power supply port and the offshore power exchange platform, respectively. It indicates the operation time of the unmanned ship A after it joins the communication channel from the power supply port. It represents the operation of converting the distance traveled by the unmanned ship A in the communication channel into the coordinates of the unmanned ship A based on the coordinates of the communication channel;

[0089] In the operation time period, a time period is At any moment, it is determined that the union of the intervals covered by the wireless signals of N unmanned ships A just covers the communication channel, which means that at least N unmanned ships A are needed to build the communication channel.

[0090] It should be noted that the present invention does not exclude setting more than a minimum number of unmanned ships A to construct a communication channel to achieve more stable data transmission.

[0091] In one embodiment of the present invention, the operation area is divided into a plurality of sub-areas, and the number of unmanned ships B and unmanned ships C in the operation area is determined according to the number of sub-areas, including:

[0092] Divide the operation area into several sub-areas and assign an unmanned ship C to each sub-area;

[0093] The maximum distance between each sub-region and its adjacent sub-region is obtained by brute force method, and the number of maximum distances is obtained according to the number of adjacent sub-regions;

[0094] Based on the ultrashort wave propagation distance combined with K-means clustering of the maximum distance, several signal missing areas are obtained according to the distribution of the maximum distance. The signal missing area is a polygon constructed by the maximum distance.

[0095] The geometric center of the signal missing area is determined by the centroid algorithm, and the unmanned ship B is gradually and iteratively filled around the geometric center of the signal missing area to ensure that the signal range of the unmanned ship B covers the signal missing area;

[0096] In a preferred embodiment of the present invention, an operation area is of any shape, and the operation area is divided into several sub-areas according to the shape of the operation area and the processing volume per unit time of the unmanned ship C; the adjacent sub-areas of each sub-area are determined, for example, if a sub-area has a boundary with four sub-areas, then the sub-area has four adjacent sub-areas, and the sub-area has a maximum distance to the four sub-areas respectively, and the two ends of the maximum distance are respectively located at the edges of the adjacent sub-areas, and the maximum distance can be calculated by the brute force method;

[0097] The edges of two adjacent sub-areas are segmented to obtain a first precision point set and a second precision point set respectively, and then the distance of the first precision point combined with each point in the second precision point set is recorded and the first maximum distance is returned, and the points of the first maximum distance are continuously segmented, recorded and the maximum distance is returned until the preset precision is reached, and the maximum distance is determined;

[0098] Combined with the sub-areas shown in the simulation diagram and the R maximum distances marked on the sub-areas, the propagation threshold is set to , and cluster analysis is performed on it through the k-means clustering algorithm to determine the maximum distance greater than the propagation threshold among the R maximum distances, and obtain several signal missing areas. The signal missing area is a polygon constructed with the maximum distance between adjacent sub-areas marked as the edge;

[0099] Determine the vertex coordinates of the signal loss area The center of the signal missing area is , the calculation formula is as follows: ; ; ;

[0100] in, represents the area of ​​the signal missing region, and n represents the number of sides of the signal missing region;

[0101] Assume that the maximum distance between two adjacent sub-regions is ;

[0102] If, one times the propagation threshold Twice the propagation threshold, the data of the adjacent sub-areas are connected by placing the unmanned ship B in the center of the signal missing area;

[0103] If, twice the propagation threshold , then through Determine the number of unmanned ships B between two adjacent sub-areas; where, It means rounding up, and the unmanned ship B is evenly distributed on the path between;

[0104] Based on the total number of unmanned ships B in the simulation diagram, the minimum number of unmanned ships B required is determined.

[0105] It should be noted that the present invention does not exclude setting more than a minimum number of unmanned ships B to connect to the unmanned ship C to achieve more stable data transmission.

[0106] In one embodiment of the present invention, after the unmanned ship C collects data from the sea area of ​​the operation area, the data is transmitted to the power supply port via wireless signals, including:

[0107] Construct a tree graph, map the offshore power exchange platform as the root node, map the unmanned ship B and the unmanned ship C as child nodes, and classify the unmanned ship B and the unmanned ship C according to the depth of the node to obtain a tree graph;

[0108] The tree diagram is updated at fixed time intervals, and the unmanned ships B and C dynamically adjust the node classification according to the depth of the corresponding nodes; the depth of the node is determined by the minimum number of hops that the node data passes through to reach the root node;

[0109] Wherein, in the tree diagram, multiple n-level nodes and multiple n-1-level nodes all establish data connection relationships, and the multiple n-level nodes are evenly distributed to multiple n-1-level nodes;

[0110] Transmit data in slices according to the tree graph and time window size at any moment;

[0111] pass Determine the amount of data transmitted by each secondary node at any time; where, Indicates the data transmission capacity of the wireless signal at any time. Indicates the number of secondary nodes connected to the root node;

[0112] pass Determine the data transmission volume of each third-level node at any time; where, Indicates the number of child nodes of any secondary node, Indicates the type of the secondary node; if the type of the secondary node is unmanned ship C, then is 1, and unmanned ship B is is 0; by looping this process, the data transmission size per unit time of each node is obtained;

[0113] At any time, if the i-th node in the tree diagram is the unmanned ship C, then determine whether the i-th node is a leaf node; if the i-th node is a leaf node, then the data slice size of the i-th node is equal to the data transmission volume of the i-th node at any time; if the i-th node is an intermediate node, then the data slice size of the i-th node is equal to the data transmission volume of the child node of the i-th node at any time;

[0114] The slice data of each time window are marked with timestamps. Unmanned ships A, B and C use timestamps as priority when transmitting, and give priority to transmitting slice data with earlier timestamps; if the timestamps of unmanned ships A, B and C are the same, the smaller slice data will be transmitted first.

[0115] In one embodiment of the present invention, the offshore power exchange platform is mapped as a root node in a tree graph, the unmanned ship B is mapped as an intermediate node in the tree graph, and the unmanned ship C is mapped as an intermediate node or a leaf node in the tree graph;

[0116] Among them, the depth of the node is determined by the minimum number of hops for the slice data of each node to be transmitted to the root node at the i-th moment; for example, the first unmanned ship C has two ways to transmit the slice data to the offshore power exchange platform; one is that the first unmanned ship C transmits it to the first unmanned ship B, and the first unmanned ship B transmits it to the offshore power exchange platform; the other is that the first unmanned ship C transmits it to the second unmanned ship C, the second unmanned ship C transmits it to the second unmanned ship B, and the second unmanned ship B transmits it to the offshore power exchange platform; the depths of the two paths are 3 and 4 respectively, then at the i-th moment, the first unmanned ship C selects a path with a depth of 3 to construct a tree graph.

[0117] In one embodiment of the present invention, at the i-th moment, based on the minimum bottleneck bandwidth in the communication channel, the data transmission capacity of the wireless signal of the communication channel at the i-th moment is determined to be the minimum bottleneck bandwidth, and the data transmission capacity of the offshore power exchange platform is also the minimum bottleneck bandwidth; in the time window executed at the i-th moment, the data transmission size per second in the time window cannot exceed the minimum bottleneck bandwidth, so as to avoid congestion in the communication channel;

[0118] According to the tree diagram at the i-th moment, the number of secondary nodes is determined to be 2, and the data transmission volume per second in the time window executed at the i-th moment of each secondary node is ; The first secondary node is the unmanned ship B, which has two child nodes. The data transmission volume per second of the two child nodes of the first secondary node in the time window of the i-th moment is ; The second secondary node is the unmanned ship C. The unmanned ship C has one child node, which is also an unmanned ship C. Since the unmanned ship C collects data itself, the data collected by itself also needs to be sliced. The data slice size of the unmanned ship C is , the data transmission volume per second of the child node unmanned ship C of the unmanned ship C in the time window at the i-th moment is , the data slice size of the child node unmanned ship C is determined according to the number and type of its child nodes; the calculation is repeated for each node in the tree diagram at the i-th moment to determine the data slice size in the data window executed by each unmanned ship C at the i-th moment; after the data slicing is completed, the sum of the slice data received by the root node per second in the time window is exactly the minimum bottleneck bandwidth.

[0119] It should be noted that the size of the time window is ≤ the interval time of tree map update.

[0120] In one embodiment of the present invention, in each time window, unmanned ships A, B and C slice the collected data respectively; when each data slice is generated, the system will mark the current timestamp for it; the timestamp reflects the time point when the data slice is generated. The earlier the timestamp, the earlier the data is generated, and the more priority is needed for transmission; during the data transmission process, unmanned ships A, B and C determine the priority according to the timestamp of the slice data; in the same time window, the transmission system will give priority to transmitting the data slice with an earlier timestamp; if the data slices generated by different unmanned ships are generated in the same time window, the system will operate according to the following rules:

[0121] Prioritize the transmission of data slices with earlier timestamps: First, unmanned ships A, B, and C transmit the data slices with the earliest timestamps, respectively, to ensure that the data with the earliest generation time can be transmitted to the offshore power exchange platform or other receiving end first;

[0122] Processing of identical timestamps: If the data slices of two or more nodes in unmanned ships A, B, and C have the same timestamp, the system will further compare the sizes of these slices and give priority to transmitting slices with smaller data volumes, thus avoiding large slice data from blocking the transmission channel.

[0123] In one embodiment of the present invention, the state information of the unmanned ship B and the unmanned ship C is collected to construct a power optimization model, including:

[0124] The power consumption, driving speed, driving direction, wind force, wind direction, water flow and water flow direction data of the unmanned boat B and the unmanned boat C are collected at fixed time intervals t; the water flow and water flow direction are measured by the monitoring units installed on the hulls of the unmanned boats B and C;

[0125] The driving speed, driving direction, wind force, wind direction, water flow and water flow direction data are normalized and constructed into training samples, and the power consumption is normalized as the sample label of the training sample; thus, a training set of the power optimization model is obtained;

[0126] The power optimization model includes an input layer and several hidden layers. The input layer is used to input the feature vector constructed after normalizing the driving speed, water flow size and water flow direction data; the hidden layer is used to output the updated state of the feature vector, and the updated state is output to the classifier. The classification space of the classifier represents the power consumption based on the driving speed, driving direction, wind force, wind direction, water flow size and water flow direction in a time interval t; the power optimization model reversely updates the weight parameters and bias parameters of the hidden layer through the mean error variance loss function, and determines the convergence of the power optimization model through cross-validation.

[0127] In one embodiment of the present invention, the calculation formula of the hidden layer of the power optimization model is as follows: ;in, represents the output of the hidden layer of the i-th layer, represents the weight parameter of the hidden layer of the i-th layer, represents the bias parameter of the hidden layer of the i-th layer, The feature vector constructed after normalization of the driving speed, water flow size and water flow direction data is passed to the i-1th layer output. express Activation function;

[0128] In one embodiment of the present invention, the calculation formula of the mean error variance loss function is as follows: ;in, represents the loss value, represents the number of samples, represents the actual power consumption of the kth sample, It represents the predicted power consumption of the kth sample by the power optimization model;

[0129] In one embodiment of the present invention, determining the convergence of the power optimization model through cross-validation includes:

[0130] Dataset division: Divide the collected data sets of unmanned boats B and C into training sets and validation sets according to a certain ratio (for example, 80% training sets and 20% validation sets); Cross-validation: Divide the training set into several subsets, select one of the subsets as the validation set each time, and use the remaining subsets to train the model, repeat K times, and ensure that each subset is used for validation; Model training and validation: Calculate the mean error variance loss value on the validation set; Record the loss value of each validation, and calculate the average loss value of all folds; Model convergence judgment: When the average loss value no longer decreases significantly after several consecutive iterations, the model is considered to have converged and training is stopped; Hyperparameter adjustment: If the model fails to converge, you can adjust the hyperparameters (such as learning rate, number of hidden layers, weight initialization, etc.) or change the model structure, and repeat the above process until the model converges.

[0131] In one embodiment of the present invention, based on the power optimization model, the unmanned ship C is scheduled to patrol and replenish power, including:

[0132] Collect the local weight of the unmanned ship C. The local weight represents the number of unmanned ships B and unmanned ships C that are disconnected from the data connection in the local network by simulating shielding the ultrashort wave signal of the unmanned ship C at any time. The unmanned ship C with a local weight of 0 and a power higher than a preset high power threshold is taken as a candidate ship, and the area where the candidate ship is located is taken as the candidate area;

[0133] The remaining power of unmanned ship B, unmanned ship C and offshore power exchange platform, and the coordinates of unmanned ship B and unmanned ship C are collected at fixed time intervals t; unmanned ship B with power below a preset low power threshold and unmanned ship C with power below a preset medium power threshold are respectively used as the first power supply ship and the second power supply ship;

[0134] Determine whether there is an overlapping unmanned ship C between the second power-supply ship and the backup ship;

[0135] If there are overlapping unmanned boats C, the overlapping unmanned boats C will directly replace part of the first power-supply boat, and the remaining unmanned boats C in the candidate boats will directly replace part of the second power-supply boat, and the replaced second power-supply boat will again replace part of the first power-supply boat;

[0136] If there is no overlapping unmanned ship C, the unmanned ship C in the candidate ship will directly replace part of the second power-supply ship, and the replaced second power-supply ship will replace part of the first power-supply ship again;

[0137] The replaced first power supply ship goes to the offshore unmanned platform to perform the task of replacing fully charged batteries. The replaced second power supply ship goes to the location of the first power supply ship to perform the data connection task. After replacing the fully charged batteries, the first power supply ship goes to the alternative area to perform patrol and monitoring tasks.

[0138] In one embodiment of the present invention, the scheduling of patrol monitoring of the unmanned ship C based on the power optimization model includes:

[0139] Collect the driving speed, driving direction, wind force, wind direction, water flow and water flow direction data of the unmanned ship B and the unmanned ship C at the i-th moment, determine the power consumption of the unmanned ship B and the unmanned ship C in the next time interval t based on the driving speed, driving direction, wind force, wind direction, water flow and water flow direction data, and obtain the remaining power of the unmanned ship B and the unmanned ship C at the i+t-th moment;

[0140] According to the wind force, wind direction, current force and current direction of unmanned ship B at the i-th moment, the number of the first power-supplying ships at the i+t-th moment is determined through the power optimization model;

[0141] Compare the number of candidate ships with the number of first replenishment ships;

[0142] If the number of candidate ships is less than or equal to the number of first power-supply ships, the power consumption of the unmanned ship C in the next time interval t is adjusted by adjusting the speed of the unmanned ship C, and the number of second power-supply ships at the i+tth moment is determined to be equal to the number of candidate ships;

[0143] If the number of candidate ships is greater than the number of first power-supply ships, the power consumption of the unmanned ship C in the next time interval t is adjusted by adjusting the speed of the unmanned ship C, and the number of second power-supply ships at the i+tth moment is determined to be equal to the number of first power-supply ships;

[0144] Among them, the driving speed of the unmanned boat C at the i-th moment is determined by combining the optimization algorithm with the power optimization model.

[0145] For example, three power thresholds of high, medium and low power levels of unmanned ships are set; at the i-th moment, it is predicted that at the i+t-th moment, there are 3 unmanned ships B below the low power threshold, 5 unmanned ships C with local weights of 0 and power levels above the high power threshold, and 1 unmanned ship C with local weights of 0 and power levels below the medium power threshold; through the scheduling of the power optimization model, at the i+t-th moment, there are 2 unmanned ships C with local weights not 0 and below the medium power threshold; at the i+t-th moment, 1 unmanned ship C with local weights of 0 and power levels below the medium power threshold replaces 1 unmanned ship B below the low power threshold, 2 unmanned ships C with local weights of 0 and power levels above the high power threshold replace 2 unmanned ships C with local weights not 0 and below the medium power threshold, and 2 unmanned ships C with local weights not 0 and below the medium power threshold replace 2 unmanned ships B below the low power threshold;

[0146] For example, at the i-th moment, it is predicted that at the i+t-th moment, there are 3 unmanned ships B with power levels below the low power threshold, and 1 unmanned ship C with a local weight of 0 and power levels above the high power threshold; through the scheduling of the power optimization model, at the i+t-th moment, there is 1 unmanned ship C with a local weight of non-0 and power levels below the medium power threshold; at the i+t-th moment, 1 unmanned ship C with a local weight of 0 and power levels above the high power threshold replaces 1 unmanned ship C with a local weight of non-0 and power levels below the medium power threshold, and 1 unmanned ship C with a local weight of non-0 and power levels below the medium power threshold replaces 1 unmanned ship B with a power level below the low power threshold;

[0147] When the unmanned ship B with a power below the low power threshold is replaced, the unmanned ship B with a power below the low power threshold goes to the offshore battery exchange platform to replace the fully charged battery. After the battery replacement is completed, the three fully charged unmanned ships B go to the alternative areas to perform cruising missions and change their identities to unmanned ships C.

[0148] In one embodiment of the present invention, the candidate ship replaces the first power-up ship or the second power-up ship, and the second power-up ship replaces the first power-up ship. The replacement score between the replacing ship and the replaced ship is calculated by the following formula: ;in, represents the replacement score between the replacing vessel and the replaced vessel, and They represent the remaining power of the replacing ship and the remaining power of the replaced ship respectively. represents the Euclidean distance between the replacing ship and the replaced ship, and Respectively represent the water flow size in the area where the replacing ship and the replaced ship are located, Indicates the cosine value of the angle between the water flow directions in the area where the replaced ship and the replaced ship are located;

[0149] The unmanned boat B with a fully charged battery selects an alternative area, and the selection scores of the unmanned boat B and the vacant alternative area are calculated by the following formula: ;in, represents the replacement score between the unmanned ship B and the vacant candidate area, represents the Euclidean distance between the offshore power exchange platform and the vacant candidate area, and They represent the water flow size in the offshore power exchange platform area and the vacant alternative area respectively. Represents the cosine value of the angle between the water flow direction of the offshore power exchange platform area and the vacant alternative area;

[0150] The vacant candidate area represents the sub-area left by the unmanned ship C whose local weight is 0 and whose power is higher than the high power threshold.

[0151] Based on the replacement scores, each replaced ship selects the replacement ship with the highest replacement score; if multiple replaced ships have the highest replacement scores corresponding to the same replacement ship, the replacement ship gives priority to the replaced ship with the highest replacement score; if multiple replaced ships have equal replacement scores corresponding to the same replacement ship, the replacement ship randomly selects one of the replaced ships; for example, there are 2 replacement ships and 2 replaced ships; the replacement scores of the first replacement ship for the 2 replaced ships are 0.80 and 0.85 respectively; the replacement scores of the second replacement ship for the 2 replaced ships are 0.75 and 0.80 respectively; thus, it is determined that the first replacement ship is assigned to the second replaced ship, and the second replacement ship is assigned to the first replaced ship.

[0152] It should be noted that the unmanned ship B only performs the data connection task, and the power consumption of this task is negligible. Therefore, the unmanned ship B has enough power to go to the offshore unmanned power exchange platform.

[0153] In one embodiment of the present invention, the speed of the unmanned ship C at the i-th moment is determined by combining the optimization algorithm with the power optimization model, including:

[0154] Randomly generate an initial particle swarm with R individuals that meet the constraints, and initialize the current number of iterations to 1;

[0155] Encode the individual, and each element of the code represents the driving speed , Driving direction , Wind force , Wind direction , water flow size and water flow direction Data; coded as follows: ; ; ;

[0156] in, and They represent the wind force and water flow size that the unmanned ship B is subjected to, and They represent the weight coefficient of wind force and the weight coefficient of water flow respectively, and All were obtained by linear least squares fitting;

[0157] The constraints are that the driving speed is not 0, the driving direction, wind direction and water flow direction do not change, and the wind force and water flow size satisfy the linear relationship with the driving speed and driving direction;

[0158] The fitness function of each individual is obtained through the power optimization model. The fitness function value represents the power consumption in the next time interval t. The calculation formula is as follows: in represents the output of the power optimization model for individual codes, Indicates driving speed The weight coefficient of

[0159] Sort the R individuals according to their fitness values, retain a fixed number of individuals, and update the remaining individuals as parents;

[0160] Among them, only the driving speed of the parent individual is updated, and the wind force and water flow are updated according to the updated driving speed;

[0161] Determine whether the current number of iterations is equal to the maximum number of iterations; if the current number of iterations = the maximum number of iterations, determine the speed of the unmanned ship C at the i-th moment , otherwise continue iterating.

[0162] In one embodiment of the present invention, 100 individuals are randomly generated, and each individual is encoded according to the constraint conditions; then, the fitness value of each individual is output through the power optimization model, and the 100 individuals are sorted from large to small according to the fitness value; the fitness value is calculated as follows: ;in, represents the output of the power optimization model for individual codes, The weight coefficient representing the driving speed;

[0163] Keep the 10 individuals with the smallest fitness values, and update the driving speeds of the remaining 90 individuals. The update formula is as follows: ; in, represents the updated driving speed, Indicates the current driving speed. , and represent the first weight parameter, the second weight parameter and the third weight parameter respectively, represents the current individual optimal speed, represents the global individual optimal speed, and denote the first disturbance parameter and the second disturbance parameter respectively, , .

[0164] According to the updated driving speed, the corresponding code in the individual is updated; after the individual update is completed, the fitness value of each individual is output again through the power optimization model; the process is repeated until the number of iterations reaches the set maximum number of iterations;

[0165] When the set maximum number of iterations is reached, the fitness value of each individual is output again through the power optimization model, and the 100 individuals are sorted from large to small according to the fitness value;

[0166] For example, , where when the fitness value is 0.6, the power consumption is 1.8, which just makes the power consumption of the corresponding unmanned boat C at the i+tth moment equal to the medium power threshold, and the driving speed of the unmanned boat C is determined based on the driving speed in the individual code.

[0167] In one embodiment of the present invention, unmanned ship B and unmanned ship C determine their unmanned ship types based on the tasks to be performed, and respond to the dispatch based on the types of the unmanned ships.

[0168] It should be noted that when unmanned ship B and unmanned ship C are on the way to replace the corresponding unmanned ship, they do not respond to other dispatches.

[0169] For example, when an unmanned boat C in the second power-supporting boat replaces an unmanned boat B in the first power-supporting boat, the unmanned boat C does not respond to any other dispatches on the way to the unmanned boat B, such as being divided into a backup boat or the second power-supporting boat; when the unmanned boat C arrives at the location of the unmanned boat B, the unmanned boat C changes its identity to the unmanned boat B; when the unmanned boat B arrives at the backup area for patrol and monitoring tasks after replacing the fully charged battery, it changes its identity to the unmanned boat C.

[0170] The present invention provides an unmanned ship self-organizing power supply system, comprising:

[0171] The local area network module is used to establish the communication connection between the power supply port, the unmanned power exchange platform and the operation area. The data transmission module is used to segment and timestamp the data collected by the unmanned ship C and transmit it according to the priority. The scheduling module is used to schedule the power supply of the unmanned ship C and the unmanned ship B, and adjust the power consumption of the unmanned ship C through the preset power optimization model.

[0172] The above describes an embodiment of the present embodiment, but the present embodiment is not limited to the above-mentioned specific implementation methods. The above-mentioned specific implementation methods are merely illustrative and not restrictive. Under the guidance of the present embodiment, ordinary technicians in this field can also make many forms, all of which are within the protection of the present embodiment.

Claims

1. A self-organized power supply method for an unmanned ship, characterized in that: include: Step S101, setting a communication channel between the power supply port and the cruising area, and determining the number of unmanned ships A according to the set communication channel; Unmanned ship A is used to obtain fully charged batteries from the power supply port and transport them to the offshore power exchange platform, and maintain the data connection between the offshore power exchange platform and the power supply port; Among them, the offshore power exchange platform and the power supply port are respectively used as the two ends of the communication channel. The offshore power exchange platform is an offshore floating platform at an anchored position. Both the offshore power exchange platform and the power supply port are equipped with equipment for automatic power exchange for unmanned ships. Step S102, dividing the operation area into several sub-areas, and determining the number of unmanned ships B and unmanned ships C in the operation area according to the number of sub-areas; unmanned ship B is used to receive fully charged batteries from the offshore power exchange platform and maintain data connection between unmanned ship C and the offshore power exchange platform; unmanned ship C is used to perform patrol monitoring tasks; Step S103, collecting status information of the unmanned boat B and the unmanned boat C, and constructing a power optimization model; wherein the status information includes: power consumption, driving speed, driving direction, wind force, wind force direction, water flow force and water flow direction data; Step S104: Based on the power optimization model, the unmanned boat C is patrolled and monitored and the power supply is scheduled.

2. The method for self-organizing power supply of an unmanned ship according to claim 1, characterized in that: According to the set communication channel, determine the number of unmanned ships A, including: Collect the length of the communication channel And battery replacement operation time ; Among them, the length of the communication channel Indicates the length of the path set between the power supply port and the offshore power exchange platform; Determine the maximum moving speed of unmanned ship A is , moving speed represents the maximum value of the intersection of the speed ranges of all unmanned ships A; pass Determine the round trip time of unmanned ship A in the communication channel ; Based on round trip time Determine the unmanned ship A at any time during the operation period The coordinate position is calculated as follows: When the unmanned ship A moves, the coordinates ; When the unmanned ship A is stationary, the coordinates ;in, and They represent the time when the unmanned ship A arrives at the power supply port and the offshore power exchange platform, respectively. It indicates the operation time of the unmanned ship A after it joins the communication channel from the power supply port. It represents the operation of converting the distance traveled by the unmanned ship A in the communication channel into the coordinates of the unmanned ship A based on the coordinates of the communication channel; pass Determine the signal coverage of unmanned ship A ; Wherein, h represents the maximum antenna height of the unmanned ship A; based on the coordinates of the unmanned ship A at any time and the wireless signal coverage range, the coverage range of the unmanned ship A at any time during the operation period is obtained; the number of unmanned ships A is gradually increased iteratively, and the number of unmanned ships A is determined to be N; In the operation time period, a time period is At any time, determine the union of the wireless signal coverage intervals of N unmanned ships A to cover the communication channel.

3. The method for self-organizing power supply of an unmanned ship according to claim 1, characterized in that: The operation area is divided into several sub-areas, and the number of unmanned ships B and C in the operation area is determined according to the number of sub-areas, including: Divide the operation area into several sub-areas and assign an unmanned ship C to each sub-area; The maximum distance between each sub-region and its adjacent sub-region is obtained by brute force method, and the number of maximum distances is obtained according to the number of adjacent sub-regions; Based on the ultrashort wave propagation distance combined with K-means clustering of the maximum distance, several signal missing areas are obtained according to the distribution of the maximum distance. The signal missing area is a polygon constructed by the maximum distance. The geometric center of the signal missing area is determined by the centroid algorithm, and the unmanned ship B is gradually and iteratively filled around the geometric center of the signal missing area to ensure that the signal range of the unmanned ship B covers the signal missing area.

4. The method for self-organizing power supply of an unmanned ship according to claim 1, characterized in that: After the unmanned ship C collects data from the sea area of ​​the operation area, it transmits the data to the power supply port through wireless signals, including: Construct a tree graph, map the offshore power exchange platform as the root node, map the unmanned ship B and the unmanned ship C as child nodes, and classify the unmanned ship B and the unmanned ship C according to the depth of the node to obtain a tree graph; The tree diagram is updated at fixed time intervals, and the unmanned ships B and C dynamically adjust the node classification according to the depth of the corresponding nodes; the depth of the node is determined by the minimum number of hops that the node data passes through to reach the root node; Wherein, in the tree diagram, multiple n-level nodes and multiple n-1-level nodes all establish data connection relationships, and the multiple n-level nodes are evenly distributed to multiple n-1-level nodes; Transmit data in slices according to the tree graph and time window size at any moment; pass Determine the amount of data transmitted by each secondary node at any time; where, Indicates the data transmission capacity of the wireless signal at any time. The number of secondary nodes connected to the root node determines the amount of data transmitted by each secondary node at any time; Indicates the data transmission capacity of the wireless signal at any time. Indicates the number of secondary nodes connected to the root node; pass Determine the data transmission volume of each third-level node at any time; where, Indicates the number of child nodes of any secondary node, Indicates the type of the secondary node; if the type of the secondary node is unmanned ship C, then is 1, and unmanned ship B is is 0; by looping this process, the data transmission size per unit time of each node is obtained; At any time, if the i-th node in the tree diagram is the unmanned ship C, then determine whether the i-th node is a leaf node; if the i-th node is a leaf node, then the data slice size of the i-th node is equal to the data transmission volume of the i-th node at any time; if the i-th node is an intermediate node, then the data slice size of the i-th node is equal to the data transmission volume of the child node of the i-th node at any time; The slice data of each time window are marked with timestamps. Unmanned ships A, B and C use timestamps as priority when transmitting, and give priority to transmitting slice data with earlier timestamps; if the timestamps of unmanned ships A, B and C are the same, the smaller slice data will be transmitted first.

5. The method for self-organizing power supply of an unmanned ship according to claim 1, characterized in that: A method for self-organizing power supply of unmanned ships, characterized in that the status information of unmanned ships B and C is collected to construct a power optimization model, including: The power consumption, driving speed, driving direction, wind force, wind direction, water flow and water flow direction data of the unmanned boat B and the unmanned boat C are collected at fixed time intervals t; the water flow and water flow direction are measured by the monitoring units installed on the hulls of the unmanned boats B and C; The driving speed, driving direction, wind force, wind direction, water flow and water flow direction data are normalized and constructed into training samples, and the power consumption is normalized as the sample label of the training sample; thus, a training set of the power optimization model is obtained; The power optimization model includes an input layer and several hidden layers. The input layer is used to input the feature vector constructed after normalizing the driving speed, water flow size and water flow direction data; the hidden layer is used to output the updated state of the feature vector, and the updated state is output to the classifier. The classification space of the classifier represents the power consumption based on the driving speed, driving direction, wind force, wind direction, water flow size and water flow direction in a time interval t; the power optimization model reversely updates the weight parameters and bias parameters of the hidden layer through the mean error variance loss function, and determines the convergence of the power optimization model through cross-validation.

6. The method for self-organizing power supply of an unmanned ship according to claim 5, characterized in that: Based on the power optimization model, the patrol and power replenishment scheduling of the unmanned ship C is carried out, including: Collect the local weight of the unmanned ship C. The local weight represents the number of unmanned ships B and unmanned ships C that are disconnected from the data connection in the local network by simulating shielding the ultrashort wave signal of the unmanned ship C at any time. The unmanned ship C with a local weight of 0 and a power higher than a preset high power threshold is taken as a candidate ship, and the area where the candidate ship is located is taken as the candidate area; The remaining power of unmanned ship B, unmanned ship C and offshore power exchange platform, and the coordinates of unmanned ship B and unmanned ship C are collected at fixed time intervals t; unmanned ship B with power below a preset low power threshold and unmanned ship C with power below a preset medium power threshold are respectively used as the first power supply ship and the second power supply ship; Determine whether there is an overlapping unmanned ship C between the second power-supply ship and the backup ship; If there are overlapping unmanned boats C, the overlapping unmanned boats C will directly replace part of the first power-supply boat, and the remaining unmanned boats C in the candidate boats will directly replace part of the second power-supply boat, and the replaced second power-supply boat will again replace part of the first power-supply boat; If there is no overlapping unmanned ship C, the unmanned ship C in the candidate ship will directly replace part of the second power-supply ship, and the replaced second power-supply ship will replace part of the first power-supply ship again; The replaced first power supply ship goes to the offshore unmanned platform to perform the task of replacing fully charged batteries. The replaced second power supply ship goes to the location of the first power supply ship to perform the data connection task. After replacing the fully charged batteries, the first power supply ship goes to the alternative area to perform patrol and monitoring tasks.

7. The method for self-organizing power supply of an unmanned ship according to claim 6, characterized in that: Based on the power optimization model, the unmanned ship C is dispatched for patrol monitoring, including: Collect the driving speed, driving direction, wind force, wind direction, water flow and water flow direction data of the unmanned ship B and the unmanned ship C at the i-th moment, determine the power consumption of the unmanned ship B and the unmanned ship C in the next time interval t based on the driving speed, driving direction, wind force, wind direction, water flow and water flow direction data, and obtain the remaining power of the unmanned ship B and the unmanned ship C at the i+t-th moment; According to the wind force, wind direction, current force and current direction of unmanned ship B at the i-th moment, the number of the first power-supplying ships at the i+t-th moment is determined through the power optimization model; Compare the number of candidate ships with the number of first replenishment ships; If the number of candidate ships is less than or equal to the number of first power-supply ships, the power consumption of the unmanned ship C in the next time interval t is adjusted by adjusting the speed of the unmanned ship C, and the number of second power-supply ships at the i+tth moment is determined to be equal to the number of candidate ships; If the number of candidate ships is greater than the number of first power-supply ships, the power consumption of the unmanned ship C in the next time interval t is adjusted by adjusting the speed of the unmanned ship C, and the number of second power-supply ships at the i+tth moment is determined to be equal to the number of first power-supply ships; Among them, the driving speed of the unmanned boat C at the i-th moment is determined by combining the optimization algorithm with the power optimization model.

8. The method for self-organizing power supply of an unmanned ship according to claim 7, characterized in that: The driving speed of the unmanned ship C at the i-th moment is determined by combining the optimization algorithm with the power optimization model, including: Randomly generate an initial particle swarm with R individuals that meet the constraints, and initialize the current number of iterations to 1; Encode the individual, and each element of the code represents the driving speed , Driving direction , Wind force , Wind direction , water flow size and water flow direction Data; coded as follows: ; ; ; in, and They represent the wind force and water flow size that the unmanned ship B is subjected to, and They represent the weight coefficient of wind force and the weight coefficient of water flow respectively, and All were obtained by linear least squares fitting; The constraints are that the driving speed is not 0, the driving direction, wind direction and water flow direction do not change, and the wind force and water flow size satisfy the linear relationship with the driving speed and driving direction; The fitness function of each individual is obtained through the power optimization model. The fitness function value represents the power consumption in the next time interval t. The calculation formula is as follows: ;in represents the output of the power optimization model for individual codes, Indicates driving speed The weight coefficient of Sort the R individuals according to their fitness values, retain a fixed number of individuals, and update the remaining individuals as parents; Among them, only the driving speed of the parent individual is updated, and the wind force and water flow are updated according to the updated driving speed; Determine whether the current number of iterations is equal to the maximum number of iterations; if the current number of iterations = the maximum number of iterations, determine the speed of the unmanned ship C at the i-th moment , otherwise continue iterating.

9. A method for self-organizing power supply for an unmanned ship according to any one of claims 1 to 7, characterized in that: Unmanned ship B and unmanned ship C determine their unmanned ship types based on the missions they are performing, and respond to scheduling based on the types of unmanned ships.

10. An unmanned ship self-organizing power supply system, characterized in that: Applied to the method according to any one of claims 1 to 9, comprising: Local area network module, used to build communication connections between the power supply port, unmanned power exchange platform and operation area; The data transmission module is used to segment and timestamp the data collected by the unmanned ship C and transmit them according to the priority; The scheduling module is used to perform power replenishment scheduling for unmanned ship C and unmanned ship B, and adjust the power consumption of unmanned ship C through a preset power optimization model.

Citation Information

Patent Citations

  • Full-power ship cluster path optimization method considering power conversion and shore power access

    CN114912696A

  • Electric energy processing method and related device

    CN116707108A

  • Automatic power conversion method, system and equipment for ship and storage medium

    CN116788106A

  • Charging scheduling method, device and system for electric ship

    CN116799881A

  • Hybrid power ship energy management method and system considering energy storage safety and shore power supply

    CN118367637A

Cited By

  • Dynamic electric quantity balance and collaborative replenishment system among fleets

    CN120278482A

  • Unmanned ship task planning and fuel supply cooperative scheduling system

    CN121560034A

  • Whole-process collaborative scheduling management system for autonomous refueling operation of unmanned ship

    CN121616062A

  • Whole-process collaborative scheduling management system for unmanned ship autonomous refueling operation

    CN121616062B