Unmanned boat formation method and system based on Beidou satellite navigation system

Abstract

The invention belongs to the technical field of intelligent ship navigation and positioning, and discloses an unmanned ship formation method and system based on a Beidou satellite navigation system. According to the method, a mobile Beidou master station installed on a first formation unmanned ship, mobile Beidou slave stations on other formation unmanned ships, and inertial navigation devices, MESH communication devices and display controllers on the formation ships are included. And the mobile Beidou slave station receives the correction data of the mobile Beidou master station to realize RTK relative high-precision positioning between the formation boats. The Beidou equipment on each boat and inertial navigation equipment on the boat can realize combined navigation, so that the real-time navigation positioning precision is improved; and high-precision information such as position, speed, course, heading, attitude, time reference and the like is provided for a display control system, a navigation control system, a sensing system and other systems needing navigation positioning time service information of each formation unmanned system.

Description

Technical Field

[0001] The present invention belongs to the technical field of intelligent boat navigation and positioning, and in particular relates to an unmanned boat formation method and system based on a Beidou satellite navigation system. Background Art

[0002] Although traditional unmanned boats are equipped with satellite and inertial navigation equipment, their navigation positioning error reaches 3 to 10 meters, which cannot meet the control requirements of high-precision unmanned boat formations. Therefore, in order to improve the control accuracy of unmanned boat formations to meet more extreme usage requirements, it is necessary to design a high-precision unmanned boat formation control method and system.

[0003] RTK differential positioning technology is a high-precision positioning and navigation technology based on satellites, with positioning accuracy reaching the centimeter level. However, traditional RTK navigation and positioning technologies are all based on RTK navigation and positioning of satellite stations, Qianxun and fixed base stations. Satellite station RTK navigation and positioning can broadcast differential data to the world through satellites, so that the receiving equipment can achieve RTK navigation and positioning. The disadvantage is that the cost continues to increase with the increase of service life. Although Qianxun and fixed base stations have high accuracy and low cost, they require the use of the ground-based enhanced inertial navigation network formed by 3G / 4G / 5G communication technology, which is exactly what marine vessels do not have. Summary of the invention

[0004] In order to overcome the problems existing in the prior art, the disclosed embodiments of the present invention provide an unmanned boat formation method and system based on the Beidou satellite navigation system. Specifically, it relates to a high-precision navigation, positioning and timing method for an unmanned boat formation based on the Beidou satellite navigation system. The purpose of the present invention is to improve the navigation and positioning accuracy of unmanned boats and realize RTK relatively high-precision positioning between formation boats.

[0005] The technical solution is as follows: An unmanned boat formation method based on the Beidou satellite navigation system includes: S1, select one of the unmanned boats in the formation, define it as the first unmanned boat, install the mobile Beidou master station and antenna, and connect the communication interface; S2, the rest of the fleet Install the mobile BeiDou slave station and antenna, , Indicate the total number of unmanned boats in the formation and connect the communication interface; S3, the mobile Beidou master station periodically outputs the single-point navigation positioning timing data and differential second pulse information to the first inertial navigation device; the first inertial navigation device periodically transmits the combined navigation positioning, speed, heading, heading, attitude and timing information; S4, the mobile Beidou master station transparently transmits the received data to the first MESH device, and transmits it to the MESH device via wireless communication , MESH devices Transmit the received differential correction data to the display controller , through the display controller Transmitted to the mobile Beidou slave station ; S5, Mobile Beidou Slave Station After receiving the differential correction data from the mobile Beidou master station, it periodically outputs the RTK navigation positioning timing data and differential second pulse information to the inertial navigation device ; Inertial navigation equipment Periodically transmit the combined navigation positioning, speed, heading, heading, attitude and timing information; S6, Mobile Beidou Slave Station The differential correction data transmitted by the mobile Beidou master station is used to determine whether the communication of the MESH device is interrupted, and to transmit the navigation positioning, speed, heading, heading, attitude and timing information.

[0006] In step S1, communication interface docking is performed, including: First, connect the differential correction data communication interface 422 of the mobile Beidou master station to the 422 serial port of the first display controller; Then connect the serial port 422 of the navigation and positioning data and the differential second pulse interface to the corresponding interface of the first inertial navigation device; Finally, the navigation, positioning and timing data of the inertial navigation device are connected to the first display controller, the first navigation control system, the first perception system and the first system of navigation, positioning and timing information through the 422 serial port; The first display controller is connected to the first MESH device via the LAN port; In step S2, docking the communication interface includes: First move the Beidou slave Differential correction data communication 422 interface and display controller 422 serial port connection; Then connect the navigation positioning data 422 serial port and differential second pulse interface to the inertial navigation device The corresponding interface connection; Finally, the navigation positioning and timing data of the inertial navigation device are transmitted to the first Display controller, Navigation control system, Perception system and navigation positioning and timing information System docking; No. The display controller communicates with the MESH device connection.

[0007] In step S3, the mobile Beidou master station periodically outputs the single-point navigation positioning timing data and differential second pulse information to the first inertial navigation device, including: the mobile Beidou master station outputs the single-point navigation positioning timing data and differential second pulse information to the first inertial navigation device through the 422 serial port and the differential second pulse interface at a period of 1 Hz; the first inertial navigation device transmits the combined navigation positioning, speed, heading, heading, attitude and timing information to the first display controller, the first navigation control system, the first perception system and the first system of navigation positioning timing information through the 422 serial port at a period of 100 Hz; In step S4, the mobile Beidou master station transparently transmits the received data to the first MESH device, including: the mobile Beidou master station outputs the differential correction data to the first display controller through the 422 serial port, and after the first display controller receives the differential correction data, it transparently transmits the received data to the first MESH device through the LAN port.

[0008] In step S5, the Beidou slave station is moved After receiving the differential correction data from the mobile Beidou master station, the RTK navigation positioning timing data and differential second pulse information are output to the first Inertial navigation equipment; The inertial navigation device transmits the integrated navigation positioning, speed, heading, heading, attitude and timing information to the first Display controller, Navigation control system, Perception system and navigation positioning and timing information system.

[0009] In step S5, the inertial navigation device Periodically transmit the combined navigation positioning, speed, heading, heading, attitude and timing information, including: Step 1: By representing and learning the inertial navigation network reconstructed by vectors of different nodes of navigation positioning, speed, heading, heading, attitude and timing information, extract some regional structures of different nodes of navigation positioning, speed, heading, heading, attitude and timing information in the inertial navigation network, and integrate them into a low-dimensional, dense vector space, and obtain the vector representation of different nodes of navigation positioning, speed, heading, heading, attitude and timing information in the inertial navigation network; Step 2: using the vector representation of different nodes of navigation positioning, speed, heading, heading, attitude and timing information in the inertial navigation network and the marking information of different nodes of some navigation positioning, speed, heading, heading, attitude and timing information, the high-order meaning information of different nodes of some navigation positioning, speed, heading, heading, attitude and timing information in the inertial navigation network structure is integrated into the inertial navigation network structure to enhance the topological structure of the inertial navigation network; Step three, using the graph reconstructed based on the vectors of different nodes of navigation positioning, speed, heading, heading, attitude and timing information to integrate the enhanced partial regional structure of different nodes of navigation positioning, speed, heading, heading, attitude and timing information in the inertial navigation network, and integrating some high-order meaning information of different nodes of marked navigation positioning, speed, heading, heading, attitude and timing information into the vector representation of different nodes of navigation positioning, speed, heading, heading, attitude and timing information; Step four, use the fusion learning method to fuse the vector representations of different nodes of navigation positioning, speed, heading, heading, attitude and timing information, so as to realize the fusion of different nodes of navigation positioning, speed, heading, heading, attitude and timing information.

[0010] In step 3, the graph reconstructed based on the vectors of different nodes of navigation positioning, speed, heading, heading, attitude and timing information is integrated into the enhanced partial regional structure of different nodes of navigation positioning, speed, heading, heading, attitude and timing information in the inertial navigation network, including: (1) Calculate the degree of integration of partial regional structural information of different nodes of navigation positioning, speed, heading, heading, attitude and timing information into different node vectors of navigation positioning, speed, heading, heading, attitude and timing information; The partial regional structure of different nodes of navigation positioning, speed, heading, heading, attitude and timing information refers to the relationship between different nodes of navigation positioning, speed, heading, heading, attitude and timing information and different nodes of navigation positioning, speed, heading, heading, attitude and timing information in their neighborhood. In the vector space, the similarity between vectors is used to represent the proximity between different nodes of corresponding navigation positioning, speed, heading, heading, attitude and timing information. The closer the navigation positioning, speed, heading, heading, attitude and timing information are in the inertial navigation network, the closer the navigation positioning, speed, heading, heading, attitude and timing information are. The more similar the vector representations of different pairs of navigation positioning, speed, heading, heading, attitude and timing information nodes are in the vector space, the more dissimilar the vector representations of different pairs of navigation positioning, speed, heading, heading, attitude and timing information nodes are in the vector space. From this perspective, a method is proposed to calculate the degree of information integration of different parts of the navigation positioning, speed, heading, heading, attitude and timing information nodes in the vectors of different nodes in the inertial navigation network. set up Represents a node in the inertial navigation network that contains different information on navigation positioning, speed, heading, heading, attitude and timing. Indicates different nodes of navigation positioning, speed, heading, heading, attitude and timing information The vector representation of Indicates different nodes of navigation positioning, speed, heading, heading, attitude and timing information A collection of vector representations of the first-order neighbors of the object; Analyze the navigation positioning, speed, heading, heading, attitude and timing information of different nodes The first-order neighborhood of the first-order neighbor obtains the first-order neighbor navigation positioning, speed, heading, heading, attitude and timing information of different nodes vector set , and Considered as different nodes of navigation positioning, speed, heading, heading, attitude and timing information Positive sample set to calculate and The average cosine similarity between : In the formula, the symbol represents the vector inner product, It represents the second normal form; Select from the inertial navigation network Unconnected The navigation positioning, speed, heading, heading, attitude and timing information of different nodes are represented as negative sample sets. , and calculate With their mean cosine similarity , the expression is: By calculation and The difference between the different nodes is used to obtain navigation positioning, speed, heading, heading, attitude and timing information. Integration The degree of structural information of different nodes in the navigation positioning, speed, heading, heading, attitude and timing information contained in the , the expression is:

[0011] (2) Reconstruct the vector representation of different nodes of navigation positioning, speed, heading, heading, attitude and timing information; Construct a reconstruction operator of vector representations of different nodes of navigation positioning, speed, heading, heading, attitude and timing information, and reconstruct the vector representations of different nodes of navigation positioning, speed, heading, heading, attitude and timing information by using the fragments of the vector representations of different nodes of the first-order neighborhood navigation positioning, speed, heading, heading, attitude and timing information at each iteration, so as to improve the degree of integration of partial regional structures of different nodes of navigation positioning, speed, heading, heading, attitude and timing information in the vectors of different nodes of navigation positioning, speed, heading, heading, attitude and timing information; set up Reconstruct the navigation positioning, speed, heading, heading, attitude and timing information of different nodes in the inertial navigation network When the vector is expressed as Different sets of neighboring nodes with navigation positioning, speed, heading, heading, attitude and timing information In the navigation positioning, speed, heading, heading, attitude and timing information of different nodes The values ​​are sampled in a roulette manner The ordered set obtained by sampling different nodes of navigation positioning, speed, heading, heading, attitude and timing information is recorded as , The The different nodes of navigation positioning, speed, heading, heading, attitude and timing information are recorded as ;Will The vector representation of each node of navigation positioning, speed, heading, heading, attitude and timing information is divided into segments of equal length, The first set of fragments of vector representations of different nodes of navigation positioning, speed, heading, heading, attitude and timing information is expressed as: In the formula, Different nodes for navigation positioning, speed, heading, heading, attitude and timing information Middle The vector representation of different nodes of navigation positioning, speed, heading, heading, attitude and timing information fragments; The navigation positioning, speed, heading, heading, attitude and timing information of different nodes are represented by the corresponding vectors. The reconstruction operators are used to reconstruct the values ​​in sequence from small to large. The vector representations of different nodes of navigation positioning, speed, heading, heading, attitude and timing information with a low degree of integration of some regional structural information are reconstructed first, and the vector representations of different nodes of navigation positioning, speed, heading, heading, attitude and timing information with a high degree of integration are retained, so that the vector representations of different nodes of navigation positioning, speed, heading, heading, attitude and timing information with a high degree of integration are made neighbors of the vector representations of different nodes of navigation positioning, speed, heading, heading, attitude and timing information with a low degree of integration, and have more opportunities to participate in the reconstruction process, so that the fragments of the vector representations of different nodes of navigation positioning, speed, heading, heading, attitude and timing information with a high degree of integration are propagated to a wider range.

[0012] In step 2, the vector representation of different nodes of navigation positioning, speed, heading, heading, attitude and timing information in the inertial navigation network and the marking information of different nodes of some navigation positioning, speed, heading, heading, attitude and timing information are used, including: the different nodes of navigation positioning, speed, heading, heading, attitude and timing information in the inertial navigation network are divided into two subsets according to whether they are marked or not. and , Represents different sets of nodes with marked navigation position, speed, heading, heading, attitude and timing information. Represents different sets of nodes of unlabeled navigation position, speed, heading, heading, attitude and timing information. The navigation positioning, speed, heading, heading, attitude and timing information in the different nodes are divided into different categories according to their categories. subsets, It represents the total number of different node categories of navigation positioning, speed, heading, heading, attitude and timing information, denoted as ,Will Navigation positioning, speed, heading, heading, attitude and timing information in different nodes The corresponding mark is , The set of different nodes with marked first-order neighbor navigation positioning, speed, heading, heading, attitude and timing information is recorded as The inertial navigation network structure is enhanced by adding a certain number of edges between different nodes of navigation positioning, speed, heading, heading, attitude and timing information that are marked and belong to the same category. Different nodes of navigation positioning, speed, heading, heading, attitude and timing information Count the number of edges added with it as one end , the expression is: In the formula, Indicates different nodes of navigation positioning, speed, heading, heading, attitude and timing information The first-order marked points of ; express First-order marked neighborhood navigation positioning, speed, heading, heading, attitude and timing information are tracked in different nodes The preset proportion of different nodes with the same navigation positioning, speed, heading, heading, attitude and timing information, express The number of edges established when none of its first-order neighbors are labeled; Represents an equivalence judgment function.

[0013] In step four, the fusion learning method is to obtain the enhanced graph of different node structures of navigation positioning, speed, heading, heading, attitude and timing information of the attributeless inertial navigation network, and then use the machine learning fusion model to fuse the vector representations of different nodes of navigation positioning, speed, heading, heading, attitude and timing information, so as to finally achieve the purpose of fusing different nodes of navigation positioning, speed, heading, heading, attitude and timing information.

[0014] In step S6, the Beidou slave station is moved If the differential correction data transmitted by the mobile Beidou master station is not received within 2 seconds, the communication of the MESH device is considered to be interrupted; the mobile Beidou slave station Enter the single point positioning mode, and output the single point navigation positioning timing data and differential second pulse information to the first Inertial navigation equipment; The inertial navigation device transmits the integrated navigation positioning, speed, heading, heading, attitude and timing information to the first Display controller, Navigation control system, Perception system and navigation positioning and timing information system.

[0015] Another object of the present invention is to provide an unmanned boat formation system based on the Beidou satellite navigation system, which is implemented by the unmanned boat formation method based on the Beidou satellite navigation system, and the system includes: The first unmanned boat communication unit is used to select a boat in the formation, defined as the first unmanned boat, install the mobile Beidou master station and antenna, and connect the communication interface; The rest of the fleet Unit for the rest of the fleet Install the mobile BeiDou slave station and antenna, , Indicate the total number of unmanned boats in the formation and connect the communication interface; The mobile Beidou master station is used to periodically output single-point navigation positioning timing data and differential second pulse information to the first inertial navigation device; the first inertial navigation device periodically transmits the combined navigation positioning, speed, heading, heading, attitude and timing information; and transparently transmits the received data to the first MESH device, and transmits it to the MESH device through wireless communication , MESH devices Transmit the received differential correction data to the display controller , through the display controller Transmitted to the mobile Beidou slave station ; Mobile Beidou Slave Station After receiving the differential correction data from the mobile Beidou master station, it periodically outputs the RTK navigation positioning timing data and differential second pulse information to the inertial navigation device ; Inertial navigation equipment The combined navigation positioning, speed, heading, heading, attitude and timing information are transmitted periodically; as well as the differential correction data transmitted by the mobile Beidou master station, it is determined whether the MESH device communication is interrupted, and the navigation positioning, speed, heading, heading, attitude and timing information are transmitted.

[0016] In combination with all the above technical solutions, the beneficial effects of the present invention are as follows: the present invention proposes an unmanned boat formation method based on the Beidou satellite navigation system, which is composed of a mobile Beidou master station installed on the first unmanned boat in the formation, mobile Beidou slave stations on other unmanned boats in the formation, and inertial navigation equipment, MESH communication equipment and display controllers on each formation boat. The mobile Beidou slave station receives the correction data from the mobile Beidou master station to achieve RTK relatively high-precision positioning between formation boats. In addition, the Beidou equipment on each boat can also realize combined navigation with the inertial navigation equipment on the boat to improve the accuracy of real-time navigation and positioning, and provide high-precision position, speed, heading, heading, attitude, time reference and other information for the display and control system, navigation control system, perception system and other systems that need navigation positioning and timing information of each formation unmanned system.

[0017] The present invention improves the navigation and positioning accuracy of unmanned boats and unmanned boat formations to the centimeter level, and for the first time applies Beidou RTK technology to unmanned boat formations, filling the technical gap in the industry at home and abroad. Under small spacing conditions (meter level), the unmanned boat formation maintains a high-precision formation while avoiding collisions between boats in the formation, which has always been a problem that has plagued the industry. Traditional positioning and navigation methods cannot meet the requirements. The technical solution of the present invention solves this technical problem, making the navigation and positioning accuracy within the unmanned boat formation reach the centimeter level. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] The accompanying drawings herein are incorporated in and constitute a part of the specification, illustrate embodiments consistent with the present disclosure, and together with the description, serve to explain the principles of the present disclosure; Figure 1 is a flow chart of an unmanned boat formation method based on the Beidou satellite navigation system provided by an embodiment of the present invention; Figure 2 It is a schematic diagram of an unmanned boat formation method based on the Beidou satellite navigation system provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0019] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the specific embodiments of the present invention are described in detail below in conjunction with the accompanying drawings. In the following description, many specific details are set forth to facilitate a full understanding of the present invention. However, the present invention can be implemented in many other ways different from those described herein, and those skilled in the art can make similar improvements without violating the connotation of the present invention, so the present invention is not limited by the specific implementation disclosed below.

[0020] The innovation of the present invention lies in that: the present invention applies the differential positioning and navigation technology based on the Beidou system to the control of unmanned boat formations, and measures the navigation positioning, speed, heading, heading, attitude and timing information of the unmanned boats in each formation; by integrating information with inertial navigation equipment, the measurement accuracy of navigation positioning, speed, heading, heading, attitude and timing information is comprehensively improved, and the display and control system, navigation control system, perception system and navigation positioning and timing information system of each formation unmanned system are provided with position, speed, heading, heading, attitude and time reference information; the present invention applies the Beidou-based RTK high-precision differential positioning technology to the unmanned boat formation for the first time, and combines it with the inertial navigation equipment on the unmanned boat, so as to realize the high-precision positioning (centimeter level) and high-dynamic positioning data output (100hz) of the unmanned boat formation, and realizes the unmanned boat formation method and system based on the Beidou navigation system.

[0021] Embodiment 1, as Figure 1 As shown, the unmanned boat formation method based on the Beidou satellite navigation system provided by the embodiment of the present invention includes: S1, select one of the unmanned boats in the formation, define it as the first unmanned boat, install the mobile Beidou master station and antenna, and connect the communication interface; First, the mobile Beidou master station differential correction data communication 422 interface is connected to the 422 serial port of the first display controller, and then the navigation positioning data 422 serial port and differential second pulse interface are connected to the corresponding interface of the first inertial navigation device. The Beidou differential positioning data is combined with the inertial navigation device data to correct the inertial navigation positioning. Finally, the corrected high-precision navigation positioning data is output by the inertial navigation device through the 422 serial port to the first display controller, the first navigation control system, the first perception system, and other first systems that need navigation positioning timing information. The first display controller is connected to the first MESH device through the LAN port, and the differential correction data is transparently transmitted to other unmanned boats in the formation through the MESH for positioning data correction of each boat.

[0022] S2, the rest of the fleet Install the mobile BeiDou slave station and antenna, , Indicate the total number of unmanned boats in the formation and connect the communication interface; The rest of the fleet ( , Indicates the total number of unmanned boats in the formation) Install the mobile Beidou slave station and antenna. First, install the mobile Beidou slave station The differential correction data communication 422 interface is connected to the The 422 serial port of the display controller is connected, and then the 422 serial port of the navigation positioning data and the differential second pulse interface are connected to the The corresponding interface of the inertial navigation device is connected, and finally the navigation positioning and timing data of the inertial navigation device is transmitted to the first Display controller, Navigation control system, Perception systems and other systems that require navigation, positioning and timing information System Docking, The display controller is connected to the iMESH device through the LAN port.

[0023] S3, the mobile Beidou master station periodically outputs the single-point navigation positioning timing data and differential second pulse information to the first inertial navigation device; the first inertial navigation device periodically transmits the combined navigation positioning, speed, heading, heading, attitude and timing information; The mobile Beidou master station calculates the comprehensive positioning error in real time and outputs the single-point navigation positioning timing data and differential second pulse information to the first inertial navigation device through the 422 serial port and the differential second pulse interface at a 1Hz cycle. The first inertial navigation device receives the navigation timing data and differential second pulse information, and fuses them with the inertial navigation data through the KALMAN filter to calculate the precise zero position error, precise positioning cumulative error and other errors of the internal sensor of the inertial navigation device, and accurately corrects the positioning data, speed, heading, attitude, etc., and transmits the combined navigation positioning, speed, heading, heading, attitude and timing information to the first display controller, the first navigation control system, the first perception system and other first systems that need navigation positioning and timing information through the 422 serial port at a 100Hz cycle. With centimeter-level navigation and positioning data of unmanned boats, the perception system can accurately identify and relatively locate obstacles around the unmanned boats. The navigation control system can accurately control the formation and attitude of the unmanned boat formation to achieve high-precision navigation and collision avoidance. The display controller can display the precise position of the unmanned boat on the nautical chart and send the position information to other unmanned boats in the formation and the shore-based command and control system. S4, the mobile Beidou master station transparently transmits the received data to the first MESH device, and transmits it to the MESH device via wireless communication , MESH devices Transmit the received differential correction data to the display controller , through the display controller Transmitted to the mobile Beidou slave station ; The mobile Beidou master station outputs the differential correction data to the first display controller through the 422 serial port. After receiving the differential correction data, the first display controller transparently transmits the received data to the first MESH device through the LAN port (sends the data frame directly to the P address of the first MESH device), and transmits it to other iMESH devices through wireless communication ( , represents the total number of unmanned boats in the formation), The MESH device transmits the received differential correction data to the first Display controller, and The display controller transmits data to the mobile Beidou slave station through the 422 serial port ; S5, Mobile Beidou Slave Station After receiving the differential correction data from the mobile Beidou master station, it periodically outputs the RTK navigation positioning timing data and differential second pulse information to the inertial navigation device ; Inertial navigation equipment Periodically transmit the combined navigation positioning, speed, heading, heading, attitude and timing information; Mobile Beidou Slave Station ( , After receiving the differential correction data from the mobile Beidou master station, the RTK navigation positioning timing data and differential second pulse information are output to the first Inertial navigation equipment. The inertial navigation device receives navigation timing data and differential second pulse information, and fuses them with the inertial navigation data through the KALMAN filter to calculate the precise zero position error, precise positioning cumulative error and other errors of the internal sensor of the inertial navigation device, and accurately corrects the positioning data, speed, heading, attitude, etc. The combined navigation positioning, speed, heading, heading, attitude and timing information are transmitted to the first Display controller, Navigation control system, Perception systems and other systems that require navigation, positioning and timing information system; S6, Mobile Beidou Slave Station Based on the differential correction data transmitted by the mobile Beidou master station, it determines whether the communication of the MESH equipment is interrupted, and transmits navigation positioning, speed, heading, heading, attitude and timing information; Mobile Beidou Slave Station ( , If the mobile Beidou slave station fails to receive the differential correction data transmitted by the mobile Beidou master station within 2 seconds, it is determined that the communication of the MESH device is interrupted. Enter the single-point positioning mode, which is the ordinary positioning mode. The positioning accuracy of the positioning mode without differential correction cannot reach the centimeter level; the single-point navigation positioning timing data and differential second pulse information are output to the first Inertial navigation equipment. The inertial navigation device receives navigation timing data and differential second pulse information, and fuses them with the inertial navigation data through the KALMAN filter to calculate the error of the internal sensor zero position error and positioning cumulative error of the inertial navigation device, and corrects the positioning data, speed, heading, attitude, etc. The combined navigation positioning, speed, heading, heading, attitude and timing information are transmitted to the first Display controller, Navigation control system, Perception systems and other systems that require navigation, positioning and timing information system; like Figure 2As shown, as another implementation of the present invention, the principle of the unmanned boat formation method based on the Beidou satellite navigation system provided in an embodiment of the present invention.

[0024] Exemplarily, in step S5, The method of combining navigation positioning, speed, heading, heading, attitude and timing information of inertial navigation equipment includes: Step 1: extract the partial regional structure of different nodes of navigation positioning, speed, heading, heading, attitude and timing information in the inertial navigation network through the representation learning of the inertial navigation network reconstructed by the vectors of different nodes of navigation positioning, speed, heading, heading, attitude and timing information, and integrate them into the low-dimensional and dense vector space to obtain the vector representation of different nodes of navigation positioning, speed, heading, heading, attitude and timing information in the inertial navigation network; Step 2: using the vector representation of different nodes of navigation positioning, speed, heading, heading, attitude and timing information in the inertial navigation network and the marking information of different nodes of some navigation positioning, speed, heading, heading, attitude and timing information, the high-order meaning information of different nodes of some navigation positioning, speed, heading, heading, attitude and timing information in the inertial navigation network structure is integrated into the inertial navigation network structure to enhance the topological structure of the inertial navigation network; Step three, using the graph reconstructed based on the vectors of different nodes of navigation positioning, speed, heading, heading, attitude and timing information to integrate the enhanced partial regional structure of different nodes of navigation positioning, speed, heading, heading, attitude and timing information in the inertial navigation network, and integrating some high-order meaning information of different nodes of marked navigation positioning, speed, heading, heading, attitude and timing information into the vector representation of different nodes of navigation positioning, speed, heading, heading, attitude and timing information; Step four, use the fusion learning method to fuse the vector representations of different nodes of navigation positioning, speed, heading, heading, attitude and timing information, so as to realize the fusion of different nodes of navigation positioning, speed, heading, heading, attitude and timing information.

[0025] In step 1, the graph reconstruction based on different node vectors of navigation positioning, speed, heading, heading, attitude and timing information is integrated into: A graph integration model based on the reconstruction of different node vectors of navigation positioning, speed, heading, heading, attitude and timing information is constructed. Different node vector representations of navigation positioning, speed, heading, heading, attitude and timing information are used to reconstruct different node vector representations of navigation positioning, speed, heading, heading, attitude and timing information in an iterative manner, and the amount of partial regional structural information contained in different node vectors of navigation positioning, speed, heading, heading, attitude and timing information is gradually improved. The graph integration model based on different node vector reconstruction of navigation positioning, speed, heading, heading, attitude and timing information includes: (1) Calculate the degree of integration of partial regional structural information of different nodes of navigation positioning, speed, heading, heading, attitude and timing information into different node vectors of navigation positioning, speed, heading, heading, attitude and timing information; The partial regional structure of different nodes of navigation positioning, speed, heading, heading, attitude and timing information refers to the relationship between different nodes of navigation positioning, speed, heading, heading, attitude and timing information and different nodes of navigation positioning, speed, heading, heading, attitude and timing information in their neighborhood. In the vector space, the similarity between vectors is used to represent the proximity between different nodes of corresponding navigation positioning, speed, heading, heading, attitude and timing information. The closer the navigation positioning, speed, heading, heading, attitude and timing information are in the inertial navigation network, the closer the navigation positioning, speed, heading, heading, attitude and timing information are. The more similar the vector representations of different pairs of navigation positioning, speed, heading, heading, attitude and timing information nodes are in the vector space, the more dissimilar the vector representations of different pairs of navigation positioning, speed, heading, heading, attitude and timing information nodes are in the vector space. From this perspective, a method is proposed to calculate the degree of information integration of different parts of the navigation positioning, speed, heading, heading, attitude and timing information nodes in the vectors of different nodes in the inertial navigation network. set up Represents a node in the inertial navigation network that contains different information on navigation positioning, speed, heading, heading, attitude and timing. Indicates different nodes of navigation positioning, speed, heading, heading, attitude and timing information The vector representation of Indicates different nodes of navigation positioning, speed, heading, heading, attitude and timing information A collection of vector representations of the first-order neighbors of the object; Analyze the navigation positioning, speed, heading, heading, attitude and timing information of different nodes The first-order neighborhood of the first-order neighbor obtains the first-order neighbor navigation positioning, speed, heading, heading, attitude and timing information of different nodes vector set , and Considered as different nodes of navigation positioning, speed, heading, heading, attitude and timing information Positive sample set to calculate and The average cosine similarity between : In the formula, the symbol represents the vector inner product, It represents the second normal form; Select from the inertial navigation network Unconnected The navigation positioning, speed, heading, heading, attitude and timing information of different nodes are represented as negative sample sets. , and calculate With their mean cosine similarity , the expression is:

[0026] By calculation and The difference between the different nodes is used to obtain navigation positioning, speed, heading, heading, attitude and timing information. Integration The degree of structural information of different nodes in the navigation positioning, speed, heading, heading, attitude and timing information contained in the , the expression is: (2) Reconstruct the vector representation of different nodes of navigation positioning, speed, heading, heading, attitude and timing information; Construct a reconstruction operator of vector representations of different nodes of navigation positioning, speed, heading, heading, attitude and timing information, and reconstruct the vector representations of different nodes of navigation positioning, speed, heading, heading, attitude and timing information by using the fragments of the vector representations of different nodes of the first-order neighborhood navigation positioning, speed, heading, heading, attitude and timing information at each iteration, so as to improve the degree of integration of partial regional structures of different nodes of navigation positioning, speed, heading, heading, attitude and timing information in the vectors of different nodes of navigation positioning, speed, heading, heading, attitude and timing information; set up Reconstruct the navigation positioning, speed, heading, heading, attitude and timing information of different nodes in the inertial navigation network When the vector is expressed as Different sets of neighboring nodes with navigation positioning, speed, heading, heading, attitude and timing information In the navigation positioning, speed, heading, heading, attitude and timing information of different nodes The values ​​are sampled in a roulette manner The ordered set obtained by sampling different nodes of navigation positioning, speed, heading, heading, attitude and timing information is recorded as , The The different nodes of navigation positioning, speed, heading, heading, attitude and timing information are recorded as ;Will The vector representation of each node of navigation positioning, speed, heading, heading, attitude and timing information is divided into segments of equal length, The first set of fragments of vector representations of different nodes of navigation positioning, speed, heading, heading, attitude and timing information is expressed as: In the formula, Different nodes for navigation positioning, speed, heading, heading, attitude and timing information Middle The vector representation of different nodes of navigation positioning, speed, heading, heading, attitude and timing information fragments; The navigation positioning, speed, heading, heading, attitude and timing information of different nodes are represented by the corresponding vectors. The reconstruction operators are used to reconstruct the values ​​in sequence from small to large. The vector representations of different nodes of navigation positioning, speed, heading, heading, attitude and timing information with a low degree of integration of some regional structural information are reconstructed first, and the vector representations of different nodes of navigation positioning, speed, heading, heading, attitude and timing information with a high degree of integration are retained, so that the vector representations of different nodes of navigation positioning, speed, heading, heading, attitude and timing information with a high degree of integration are made neighbors of the vector representations of different nodes of navigation positioning, speed, heading, heading, attitude and timing information with a low degree of integration, and have more opportunities to participate in the reconstruction process, so that the fragments of the vector representations of different nodes of navigation positioning, speed, heading, heading, attitude and timing information with a high degree of integration are propagated to a wider range.

[0027] In step 2, the problem description of the fusion of different nodes of the inertial navigation network navigation positioning, speed, heading, heading, attitude and timing information based on graph integration includes: Given an inertial navigation network And some navigation positioning, speed, heading, heading, attitude and timing information of different nodes mark ,in, Represents different node sets of navigation positioning, speed, heading, heading, attitude and timing information of the inertial navigation network, Represents the edge set in the inertial navigation network. The problem of integrating different nodes of the inertial navigation network navigation positioning, speed, heading, heading, attitude and timing information is described as follows: The topological structure and the marking of different nodes of some navigation positioning, speed, heading, heading, attitude and timing information , mark other unmarked nodes of navigation positioning, speed, heading, heading, attitude and timing information; Build a method to efficiently represent inertial navigation network data in Euclidean space and find the mapper , the inertial navigation network Navigation positioning, speed, heading, heading, attitude and timing information in different nodes Mapping to Euclidean space dimensional vector representation , and in Contain as many different nodes as possible including navigation positioning, speed, heading, heading, attitude and timing information The relevant structural information in the inertial navigation network is obtained by splicing the vector representation of different nodes of all navigation positioning, speed, heading, heading, attitude and timing information in the inertial navigation network. Dimensional matrix ;remember Inertial Navigation Network K The integration of is a matrix No. Row vectors, corresponding to different nodes of navigation positioning, speed, heading, heading, attitude and timing information Vector representation, the graph integration process is expressed as: After different nodes of navigation positioning, speed, heading, heading, attitude and timing information are represented in Euclidean space, the semi-supervised fusion problem of different nodes of navigation positioning, speed, heading, heading, attitude and timing information becomes a traditional data fusion problem in Euclidean space. The fusion model is trained using the existing labeled data. Then, using the trained fusion Mark the different nodes of navigation positioning, speed, heading, heading, attitude and timing information that are not marked, and obtain the markings of different nodes of navigation positioning, speed, heading, heading, attitude and timing information : In step 2, the high-level meaning information of different nodes of some navigation positioning, speed, heading, heading, attitude and timing information is integrated, including: building a general inertial navigation network structure enhancement model, the specific method is as follows: The different nodes of navigation positioning, speed, heading, heading, attitude and timing information in the inertial navigation network are divided into two subsets according to whether they are marked or not and , Represents different sets of nodes with marked navigation position, speed, heading, heading, attitude and timing information. Represents different sets of nodes of unlabeled navigation position, speed, heading, heading, attitude and timing information. The navigation positioning, speed, heading, heading, attitude and timing information in the different nodes are divided into different categories according to their categories. subsets, It represents the total number of different node categories of navigation positioning, speed, heading, heading, attitude and timing information, denoted as ,Will Navigation positioning, speed, heading, heading, attitude and timing information in different nodes The corresponding mark is , The set of different nodes with marked first-order neighbor navigation positioning, speed, heading, heading, attitude and timing information is recorded as ; The inertial navigation network structure is enhanced by adding a certain number of edges between different nodes of navigation positioning, speed, heading, heading, attitude and timing information that are marked and belong to the same category. Different nodes of navigation positioning, speed, heading, heading, attitude and timing information Count the number of edges added with it as one end , the expression is: In the formula, Indicates different nodes of navigation positioning, speed, heading, heading, attitude and timing information The first-order marked neighborhood navigation positioning, speed, heading, heading, attitude and timing information are followed in different nodes The number of nodes with the same navigation position, speed, heading, heading, attitude and timing information; is a hyperparameter of the model, express First-order marked neighborhood navigation positioning, speed, heading, heading, attitude and timing information are tracked in different nodes The preset proportion of different nodes with the same navigation positioning, speed, heading, heading, attitude and timing information, express The number of edges established when none of its first-order neighbors are labeled; is a simple equivalence judgment function. If and If they are equal, Returns 1 if yes, otherwise returns 0; From A o Select from different nodes of navigation positioning, speed, heading, heading, attitude and timing information belonging to the same category Select nodes with different navigation positioning, speed, heading, heading, attitude and timing information to join the edge; Select nodes with different navigation positioning, speed, heading, heading, attitude and timing information to join the edge; Different nodes of navigation positioning, speed, heading, heading, attitude and timing information that are far away are added with edges to speed up the information propagation speed between different nodes of the same type of navigation positioning, speed, heading, heading, attitude and timing information; after learning the vector representation of different nodes of navigation positioning, speed, heading, heading, attitude and timing information in the first stage, the proximity between different nodes of navigation positioning, speed, heading, heading, attitude and timing information is integrated into the different node vectors of navigation positioning, speed, heading, heading, attitude and timing information, and the edges are used to speed up the information propagation speed between different nodes of navigation positioning, speed, heading, heading, attitude and timing information. and The similarity between the vector representations of different nodes in the navigation positioning, speed, heading, heading, attitude and timing information is expressed and The distance between different nodes of navigation positioning, speed, heading, heading, attitude and timing information; Select the one with the smallest similarity navigation positioning, speed, heading, heading, attitude and timing information of different nodes and Build connections between them to enhance The closeness between different nodes of navigation positioning, speed, heading, heading, attitude and timing information belonging to the same category, thereby accelerating the propagation speed of related marking information between different nodes of navigation positioning, speed, heading, heading, attitude and timing information; In step 4, the fusion learning method is to use the machine learning fusion model to fuse the vector representations of different nodes of navigation positioning, speed, heading, heading, attitude and timing information after obtaining the graph with enhanced structures of different nodes of navigation positioning, speed, heading, heading, attitude and timing information of the attributeless inertial navigation network, so as to finally achieve the purpose of fusing different nodes of navigation positioning, speed, heading, heading, attitude and timing information; Wherein, the fusion learning method includes: Input: Inertial Navigation Network , inertial navigation network integrated into the dimension , a tag set of different nodes of some navigation positioning, speed, heading, heading, attitude and timing information , the number of iterations in the first stage , the number of iterations in the second stage , hyperparameters ; Output: all navigation positioning, speed, heading, heading, attitude and timing information are marked at different nodes; Initialize a random Dimensional matrix , as Initial integration vector / *First stage learning* / for _ in range(iter1):Q =Model1( K, Q )Endfor K’= Model 2 / *Second stage learning* / for _ in range(iter2): Q = Model 1 ( K', Q )Endfor / * Fusion navigation positioning, speed, heading, heading, attitude and timing information different nodes represent vectors* / O all = r B ( Q, O part ); In the fusion learning method, a random initialization is performed for each node of navigation positioning, speed, heading, heading, attitude and timing information in the inertial navigation network. dimensional vector, and obtain the initial integration representation of the inertial navigation network ; Use the proposed heuristic representation learning method to integrate the structural information of the original inertial navigation network into In; using the integration matrix of the original inertial navigation network The inertial navigation network structure is enhanced by the marking information of different nodes of some navigation positioning, speed, heading, heading, attitude and timing information to obtain the inertial navigation network with enhanced structure. At the same time, the enhanced inertial navigation network is integrated with the graph based on the reconstruction of different node vectors of navigation positioning, speed, heading, heading, attitude and timing information. Some of the regional structures are integrated into In the training process, the fusion device is trained using the vector representation of different nodes of the navigation positioning, speed, heading, heading, attitude and timing information of the existing markers. After training, use The vector representations of all different nodes of navigation positioning, speed, heading, heading, attitude and timing information are fused to obtain the final fusion results of different nodes of navigation positioning, speed, heading, heading, attitude and timing information.

[0028] It can be seen from the above embodiments that the present invention improves the navigation and positioning accuracy of each boat in the unmanned boat formation through the Beidou differential positioning technology, and the navigation and positioning information is fused with the inertial navigation data through the KALMAN filter to calculate the error amounts such as the zero position error of the internal sensor of the inertial navigation device and the cumulative positioning error, and further improves the accuracy of the measurement data such as the speed, heading, and attitude of the inertial navigation device, and outputs it to other perception and intelligent control devices on the unmanned boat at a frequency of 100 Hz, thereby improving the real-time performance and improving the control performance of the unmanned boat and the formation as a whole.

[0029] The present invention extends the structure enhancement method to the problem of attribute network node fusion, makes full use of node attribute information and part of node label information to enhance the network structure, and is committed to designing a model with higher fusion performance.

[0030] Embodiment 2, the embodiment of the present invention provides an unmanned boat formation system based on the Beidou satellite navigation system, the system comprising: The first unmanned boat communication unit is used to select a boat in the formation, defined as the first unmanned boat, install the mobile Beidou master station and antenna, and connect the communication interface; The rest of the fleet Unit for the rest of the fleet Install the mobile BeiDou slave station and antenna, , Indicate the total number of unmanned boats in the formation and connect the communication interface; The mobile Beidou master station is used to periodically output single-point navigation positioning timing data and differential second pulse information to the first inertial navigation device; the first inertial navigation device periodically transmits the combined navigation positioning, speed, heading, heading, attitude and timing information; and transparently transmits the received data to the first MESH device, and transmits it to the MESH device through wireless communication , MESH devices Transmit the received differential correction data to the display controller , through the display controller Transmitted to the mobile Beidou slave station ; Mobile Beidou Slave Station After receiving the differential correction data from the mobile Beidou master station, it periodically outputs the RTK navigation positioning timing data and differential second pulse information to the inertial navigation device ; Inertial navigation equipment The combined navigation positioning, speed, heading, heading, attitude and timing information are transmitted periodically; as well as the differential correction data transmitted by the mobile Beidou master station, it is determined whether the MESH device communication is interrupted, and the navigation positioning, speed, heading, heading, attitude and timing information are transmitted.

[0031] The above description is only a preferred specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. Any modifications, equivalent substitutions and improvements made by any technician familiar with the technical field within the technical scope disclosed by the present invention and within the spirit and principles of the present invention should be covered within the protection scope of the present invention.

Claims

1. A method for unmanned boat formation based on Beidou satellite navigation system, characterized in that: The method includes: S1, select one of the unmanned boats in the formation, define it as the first unmanned boat, install the mobile Beidou master station and antenna, and connect the communication interface; S2, the rest of the fleet Install the mobile BeiDou slave station and antenna, , Indicate the total number of unmanned boats in the formation and connect the communication interface; S3, the mobile Beidou master station periodically outputs the single-point navigation positioning timing data and differential second pulse information to the first inertial navigation device; the first inertial navigation device periodically transmits the combined navigation positioning, speed, heading, heading, attitude and timing information; S4, the mobile Beidou master station transparently transmits the received data to the first MESH device, and transmits it to the MESH device via wireless communication , MESH devices Transmit the received differential correction data to the display controller , through the display controller Transmitted to the mobile Beidou slave station ; S5, Mobile Beidou Slave Station After receiving the differential correction data from the mobile Beidou master station, it periodically outputs the RTK navigation positioning timing data and differential second pulse information to the inertial navigation device ; Inertial navigation equipment Periodically transmit the combined navigation positioning, speed, heading, heading, attitude and timing information; S6, Mobile Beidou Slave Station The differential correction data transmitted by the mobile Beidou master station is used to determine whether the communication of the MESH device is interrupted, and to transmit the navigation positioning, speed, heading, heading, attitude and timing information.

2. The unmanned boat formation method based on the Beidou satellite navigation system according to claim 1 is characterized in that: In step S1, communication interface docking is performed, including: First, connect the differential correction data communication interface 422 of the mobile Beidou master station to the 422 serial port of the first display controller; Then connect the serial port 422 of the navigation and positioning data and the differential second pulse interface to the corresponding interface of the first inertial navigation device; Finally, the navigation, positioning and timing data of the inertial navigation device are connected to the first display controller, the first navigation control system, the first perception system and the first system of navigation, positioning and timing information through the 422 serial port; The first display controller is connected to the first MESH device via the LAN port; In step S2, docking the communication interface includes: First move the Beidou slave Differential correction data communication 422 interface and display controller 422 serial port connection; Then connect the navigation positioning data 422 serial port and differential second pulse interface to the inertial navigation device The corresponding interface connection; Finally, the navigation positioning and timing data of the inertial navigation device are transmitted to the first Display controller, Navigation control system, Perception system and navigation positioning and timing information System docking; No. The display controller communicates with the MESH device connection.

3. The unmanned boat formation method based on the Beidou satellite navigation system according to claim 1 is characterized in that: In step S3, the mobile Beidou master station periodically outputs the single-point navigation positioning timing data and differential second pulse information to the first inertial navigation device, including: the mobile Beidou master station outputs the single-point navigation positioning timing data and differential second pulse information to the first inertial navigation device through the 422 serial port and the differential second pulse interface at a period of 1 Hz; the first inertial navigation device transmits the combined navigation positioning, speed, heading, heading, attitude and timing information to the first display controller, the first navigation control system, the first perception system and the first system of navigation positioning timing information through the 422 serial port at a period of 100 Hz; In step S4, the mobile Beidou master station transparently transmits the received data to the first MESH device, including: the mobile Beidou master station outputs the differential correction data to the first display controller through the 422 serial port, and after the first display controller receives the differential correction data, it transparently transmits the received data to the first MESH device through the LAN port.

4. The unmanned boat formation method based on the Beidou satellite navigation system according to claim 1, characterized in that: In step S5, the Beidou slave station is moved After receiving the differential correction data from the mobile Beidou master station, the RTK navigation positioning timing data and differential second pulse information are output to the first Inertial navigation equipment; The inertial navigation device transmits the integrated navigation positioning, speed, heading, heading, attitude and timing information to the first Display controller, Navigation control system, Perception system and navigation positioning and timing information system.

5. The unmanned boat formation method based on the Beidou satellite navigation system according to claim 1, characterized in that: In step S5, the inertial navigation device Periodically transmit the combined navigation positioning, speed, heading, heading, attitude and timing information, including: Step 1: By representing and learning the inertial navigation network reconstructed by vectors of different nodes of navigation positioning, speed, heading, heading, attitude and timing information, extract some regional structures of different nodes of navigation positioning, speed, heading, heading, attitude and timing information in the inertial navigation network, and integrate them into a low-dimensional, dense vector space, and obtain the vector representation of different nodes of navigation positioning, speed, heading, heading, attitude and timing information in the inertial navigation network; Step 2: using the vector representation of different nodes of navigation positioning, speed, heading, heading, attitude and timing information in the inertial navigation network and the marking information of different nodes of some navigation positioning, speed, heading, heading, attitude and timing information, the high-order meaning information of different nodes of some navigation positioning, speed, heading, heading, attitude and timing information in the inertial navigation network structure is integrated into the inertial navigation network structure to enhance the topological structure of the inertial navigation network; Step three, using the graph reconstructed based on the vectors of different nodes of navigation positioning, speed, heading, heading, attitude and timing information to integrate the enhanced partial regional structure of different nodes of navigation positioning, speed, heading, heading, attitude and timing information in the inertial navigation network, and integrating some high-order meaning information of different nodes of marked navigation positioning, speed, heading, heading, attitude and timing information into the vector representation of different nodes of navigation positioning, speed, heading, heading, attitude and timing information; Step four, use the fusion learning method to fuse the vector representations of different nodes of navigation positioning, speed, heading, heading, attitude and timing information, so as to realize the fusion of different nodes of navigation positioning, speed, heading, heading, attitude and timing information.

6. The unmanned boat formation method based on the Beidou satellite navigation system according to claim 1, characterized in that: In step 3, the graph reconstructed based on the vectors of different nodes of navigation positioning, speed, heading, heading, attitude and timing information is integrated into the enhanced partial regional structure of different nodes of navigation positioning, speed, heading, heading, attitude and timing information in the inertial navigation network, including: (1) Calculate the degree of integration of partial regional structural information of different nodes of navigation positioning, speed, heading, heading, attitude and timing information into different node vectors of navigation positioning, speed, heading, heading, attitude and timing information; The partial regional structure of different nodes of navigation positioning, speed, heading, heading, attitude and timing information refers to the relationship between different nodes of navigation positioning, speed, heading, heading, attitude and timing information and different nodes of navigation positioning, speed, heading, heading, attitude and timing information in their neighborhood. In the vector space, the similarity between vectors is used to represent the proximity between different nodes of corresponding navigation positioning, speed, heading, heading, attitude and timing information. The closer the navigation positioning, speed, heading, heading, attitude and timing information are in the inertial navigation network, the closer the navigation positioning, speed, heading, heading, attitude and timing information are. The more similar the vector representations of different pairs of navigation positioning, speed, heading, heading, attitude and timing information nodes are in the vector space, the more dissimilar the vector representations of different pairs of navigation positioning, speed, heading, heading, attitude and timing information nodes are in the vector space. From this perspective, a method is proposed to calculate the degree of information integration of different parts of the navigation positioning, speed, heading, heading, attitude and timing information nodes in the vectors of different nodes in the inertial navigation network. set up Represents a node in the inertial navigation network that contains different information on navigation positioning, speed, heading, heading, attitude and timing. Indicates different nodes of navigation positioning, speed, heading, heading, attitude and timing information The vector representation of Indicates different nodes of navigation positioning, speed, heading, heading, attitude and timing information A collection of vector representations of the first-order neighbors of the object; Analyze the navigation positioning, speed, heading, heading, attitude and timing information of different nodes The first-order neighborhood of the first-order neighbor obtains the first-order neighbor navigation positioning, speed, heading, heading, attitude and timing information of different nodes vector set , and Considered as different nodes of navigation positioning, speed, heading, heading, attitude and timing information Positive sample set to calculate and The average cosine similarity between : In the formula, the symbol represents the vector inner product, It represents the second normal form; Select from the inertial navigation network Unconnected The navigation positioning, speed, heading, heading, attitude and timing information of different nodes are represented as negative sample sets. , and calculate With their mean cosine similarity , the expression is: By calculation and The difference between the different nodes is used to obtain navigation positioning, speed, heading, heading, attitude and timing information. Integration The degree of structural information of different nodes in the navigation positioning, speed, heading, heading, attitude and timing information contained in the , the expression is: (2) Reconstruct the vector representation of different nodes of navigation positioning, speed, heading, heading, attitude and timing information; Construct a reconstruction operator of vector representations of different nodes of navigation positioning, speed, heading, heading, attitude and timing information, and reconstruct the vector representations of different nodes of navigation positioning, speed, heading, heading, attitude and timing information by using the fragments of the vector representations of different nodes of the first-order neighborhood navigation positioning, speed, heading, heading, attitude and timing information at each iteration, so as to improve the degree of integration of partial regional structures of different nodes of navigation positioning, speed, heading, heading, attitude and timing information in the vectors of different nodes of navigation positioning, speed, heading, heading, attitude and timing information; set up Reconstruct the navigation positioning, speed, heading, heading, attitude and timing information of different nodes in the inertial navigation network When the vector is expressed as Different sets of neighboring nodes with navigation positioning, speed, heading, heading, attitude and timing information In the navigation positioning, speed, heading, heading, attitude and timing information of different nodes The values ​​are sampled in a roulette manner The ordered set obtained by sampling different nodes of navigation positioning, speed, heading, heading, attitude and timing information is recorded as , The The different nodes of navigation positioning, speed, heading, heading, attitude and timing information are recorded as ;Will The vector representation of each node of navigation positioning, speed, heading, heading, attitude and timing information is divided into segments of equal length, The first set of fragments of vector representations of different nodes of navigation positioning, speed, heading, heading, attitude and timing information is expressed as: In the formula, Different nodes for navigation positioning, speed, heading, heading, attitude and timing information Middle The vector representation of different nodes of navigation positioning, speed, heading, heading, attitude and timing information fragments; The navigation positioning, speed, heading, heading, attitude and timing information of different nodes are represented by the corresponding vectors. The reconstruction operators are used to reconstruct the values ​​in sequence from small to large. The vector representations of different nodes of navigation positioning, speed, heading, heading, attitude and timing information with a low degree of integration of some regional structural information are reconstructed first, and the vector representations of different nodes of navigation positioning, speed, heading, heading, attitude and timing information with a high degree of integration are retained, so that the vector representations of different nodes of navigation positioning, speed, heading, heading, attitude and timing information with a high degree of integration are made neighbors of the vector representations of different nodes of navigation positioning, speed, heading, heading, attitude and timing information with a low degree of integration, and have more opportunities to participate in the reconstruction process, so that the fragments of the vector representations of different nodes of navigation positioning, speed, heading, heading, attitude and timing information with a high degree of integration are propagated to a wider range.

7. The unmanned boat formation method based on the Beidou satellite navigation system according to claim 5, characterized in that: In step 2, the vector representation of different nodes of navigation positioning, speed, heading, heading, attitude and timing information in the inertial navigation network and the marking information of different nodes of partial navigation positioning, speed, heading, heading, attitude and timing information are used, including: The different nodes of navigation positioning, speed, heading, heading, attitude and timing information in the inertial navigation network are divided into two subsets according to whether they are marked or not and , Represents different sets of nodes with marked navigation position, speed, heading, heading, attitude and timing information. Represents different sets of nodes of unlabeled navigation position, speed, heading, heading, attitude and timing information. The navigation positioning, speed, heading, heading, attitude and timing information in the different nodes are divided into different categories according to their categories. subsets, It represents the total number of different node categories of navigation positioning, speed, heading, heading, attitude and timing information, denoted as ,Will Navigation positioning, speed, heading, heading, attitude and timing information in different nodes The corresponding mark is , The set of different nodes with marked first-order neighbor navigation positioning, speed, heading, heading, attitude and timing information is recorded as The inertial navigation network structure is enhanced by adding a certain number of edges between different nodes of navigation positioning, speed, heading, heading, attitude and timing information that are marked and belong to the same category. Different nodes of navigation positioning, speed, heading, heading, attitude and timing information Count the number of edges added with it as one end , the expression is: In the formula, Indicates different nodes of navigation positioning, speed, heading, heading, attitude and timing information The first-order marked points of ; express First-order marked neighborhood navigation positioning, speed, heading, heading, attitude and timing information are tracked in different nodes The preset proportion of different nodes with the same navigation positioning, speed, heading, heading, attitude and timing information, express The number of edges established when none of its first-order neighbors are labeled; Represents an equivalence judgment function.

8. The unmanned boat formation method based on the Beidou satellite navigation system according to claim 5, characterized in that: In step four, the fusion learning method is to obtain the enhanced graph of different node structures of navigation positioning, speed, heading, heading, attitude and timing information of the attributeless inertial navigation network, and then use the machine learning fusion model to fuse the vector representations of different nodes of navigation positioning, speed, heading, heading, attitude and timing information, so as to finally achieve the purpose of fusing different nodes of navigation positioning, speed, heading, heading, attitude and timing information.

9. The unmanned boat formation method based on the Beidou satellite navigation system according to claim 1, characterized in that: In step S6, the Beidou slave station is moved If the differential correction data transmitted by the mobile Beidou master station is not received within 2 seconds, the communication of the MESH device is considered to be interrupted; the mobile Beidou slave station Enter the single point positioning mode, and output the single point navigation positioning timing data and differential second pulse information to the first Inertial navigation equipment; The inertial navigation device transmits the integrated navigation positioning, speed, heading, heading, attitude and timing information to the first Display controller, Navigation control system, Perception system and navigation positioning and timing information system.

10. An unmanned boat formation system based on the Beidou satellite navigation system, characterized in that: The system is implemented by the unmanned boat formation method based on the Beidou satellite navigation system as described in any one of claims 1 to 9, and the system includes: The first unmanned boat communication unit is used to select a boat in the formation, defined as the first unmanned boat, install the mobile Beidou master station and antenna, and connect the communication interface; The rest of the fleet Unit for the rest of the fleet Install the mobile BeiDou slave station and antenna, , Indicate the total number of unmanned boats in the formation and connect the communication interface; The mobile Beidou master station is used to periodically output single-point navigation positioning timing data and differential second pulse information to the first inertial navigation device; the first inertial navigation device periodically transmits the combined navigation positioning, speed, heading, heading, attitude and timing information; and transparently transmits the received data to the first MESH device, and transmits it to the MESH device through wireless communication , MESH devices Transmit the received differential correction data to the display controller , through the display controller Transmitted to the mobile Beidou slave station ; Mobile BeiDou Slave Station After receiving the differential correction data from the mobile Beidou master station, it periodically outputs the RTK navigation positioning timing data and differential second pulse information to the inertial navigation device ; Inertial Navigation Equipment The combined navigation positioning, speed, heading, heading, attitude and timing information are transmitted periodically; as well as the differential correction data transmitted by the mobile Beidou master station, it is determined whether the MESH device communication is interrupted, and the navigation positioning, speed, heading, heading, attitude and timing information are transmitted.

Images