A method and system for controlling intelligent wireless communication of ships and boats

Abstract

The present invention relates to the field of wireless communication technology, and in particular to a method for intelligent wireless communication control applied to ships. The method comprises the following steps: using a multi-dimensional sensor to collect radar raw point cloud data, an AIS ship attribute data set, a Beidou navigation status data set, and night vision target feature data, and construct multi-source standardized pre-processed data for ships; thus, the present invention realizes functions such as multi-source heterogeneous data fusion, dynamic channel adaptive modulation, distributed relay communication guarantee, real-time dynamic route planning, and intelligent command generation, thereby solving problems existing in existing ship communication control technologies such as high data isolation, large transmission delay, rigid navigation path, slow obstacle avoidance response, and insufficient formation coordination, and significantly improves the communication stability, navigation safety, and overall combat effectiveness of ship formations in complex dynamic environments.

Description

Technical Field

[0001] The present invention relates to the field of wireless communication technology, and in particular to an intelligent wireless communication control method applied to ships and boats. Background Art

[0002] Existing intelligent wireless communication control technologies for vessels primarily rely on fixed-priority scheduling, static subcarrier allocation, and traditional link maintenance mechanisms to achieve data transmission and command issuance. However, these technologies face significant limitations in dynamic maritime formation environments. First, current communication control methods typically employ a unified subcarrier modulation strategy and lack dynamic modulation optimization for signal-to-noise ratio variations, resulting in insufficient data transmission reliability in high-noise environments. Second, existing priority queue scheduling often relies on a simple preemptive or round-robin approach, lacking a fine-grained dynamic allocation mechanism for time slot resources. This fails to effectively guarantee the priority transmission of situational data, control commands, and routine information in emergencies, and can easily lead to delays or loss of critical control commands. Third, existing technologies often rely on timeout retransmissions or static backup links to address issues such as link interruptions and increased packet loss rates in ad hoc networks. These technologies lack strategies based on real-time hop limit screening and dynamically downgraded control command generation, resulting in slow response and poor adaptability when communication links deteriorate. Furthermore, existing systems generally fail to achieve efficient integrated management of inter-vessel situational data sharing, navigation commands, and collision avoidance warning signals. Data processing is often decentralized and poorly synchronized, increasing the risk of information timeliness degradation. Finally, since most existing methods fail to deeply integrate environmental perception information with wireless communication processes, the impact of environmental changes on communication performance cannot be perceived and adjusted in a timely manner, further reducing the overall robustness of intelligent control of ship formations. Summary of the Invention

[0003] Based on this, it is necessary to provide a method for controlling intelligent wireless communication for ships and boats to solve at least one of the above technical problems.

[0004] To achieve the above object, a method for controlling intelligent wireless communication of a vessel is provided, the method comprising the following steps:

[0005] Step S1: Collect radar raw point cloud data, AIS ship attribute data set, Beidou navigation status data set and night vision target feature data through multi-dimensional sensors to construct multi-source standardized pre-processed data of ships;

[0006] Step S2: Perform three-dimensional spatial mapping on the multi-source standardized pre-processed data of the vessel and construct a target three-dimensional model; and construct an environment perception matrix based on the target three-dimensional model;

[0007] Step S3: Channel coding is performed on the environment perception matrix, and subcarrier allocation is performed to obtain a multi-carrier modulated signal; IP data packets are encapsulated according to the multi-carrier modulated signal, and relayed and transmitted by a wireless broadband ad hoc network radio station to obtain shared data on the situation of the ship formation;

[0008] Step S4: Generate a global initial route based on the shared data of the boat formation situation and construct a local obstacle avoidance path; mark the route nodes according to the local obstacle avoidance path and generate a formation collaborative navigation instruction set;

[0009] Step S5: The nautical chart is superimposed and displayed according to the formation collaborative navigation instruction set to obtain a real-time route situation map; a deviation-collision avoidance warning is performed on the real-time route situation map to obtain a three-level warning signal; a control instruction is issued to the central control system based on the three-level warning signal and the real-time route situation map to obtain a vessel wireless communication control report.

[0010] The beneficial effect of the present invention is that, through the data acquisition of multi-dimensional sensors, the integration of radar raw point cloud data, AIS ship attribute data set, Beidou navigation status data set and night vision target feature data, ships can obtain comprehensive information about their surroundings. The standardized preprocessing of these data ensures that various data sources can be subsequently processed in a unified format, providing a reliable basis for subsequent data analysis and decision-making. Through the construction of three-dimensional spatial mapping and target three-dimensional models, ships can be provided with precise spatial positioning and target recognition capabilities, thereby building an accurate environmental perception matrix, further improving the sensitivity and responsiveness of ships to changes in the surrounding environment. The process of channel coding and subcarrier allocation ensures the stability and reliability of information transmission. Through the encapsulation and relay transmission of multi-carrier modulated signals, it not only ensures the real-time situation sharing of the ship formation, but also improves the anti-interference ability and fault tolerance during data transmission. The generation of the global initial route and the construction of the local obstacle avoidance path, combined with the real-time route situation map and the yaw-collision avoidance warning, can effectively avoid the risk of collision during navigation and ensure the safe navigation of ships in complex environments. The three-level warning signal classification further optimizes navigation strategies, enabling vessels to respond promptly under varying navigation environments and conditions. By integrating these control commands into wireless communication control reports, vessels are provided with efficient navigation instructions, facilitating coordinated navigation within the fleet and ensuring high consistency and flexibility in mission execution. Overall, this series of steps enables precise navigation, intelligent obstacle avoidance, and secure communication for the fleet, optimizing navigation efficiency and safety in complex waters.

[0011] Preferably, step S1 includes the following steps:

[0012] Step S11: deploying navigation radars on the deck and mast of the vessel, operating in the X-band frequency band, with a detection range greater than 20 nautical miles and a resolution less than 30 meters, to scan the surrounding sea area, collecting target azimuth, distance, speed, and projected area data to obtain radar raw point cloud data;

[0013] Step S12: Using a one-way AIS receiver with a frequency of 161.975 MHz and a decoding rate of 1 second, the MMSI number, speed, heading, and static information of the ship are collected to generate an AIS ship attribute data set.

[0014] Step S13: Installing a visible light + infrared dual-band night vision device on the front of the boat to collect optical images in night or low-light environments, and extracting target outline, type and relative distance data to obtain night vision target feature data;

[0015] Step S14: Using the BeiDou-3 navigation terminal, the vessel's latitude and longitude, speed, and attitude angle data are collected in real time to generate a BeiDou navigation status data set;

[0016] Step S15: Check the radar point cloud data for a false alarm rate less than or equal to The sea clutter is suppressed, the timeout and unupdated signals of the AIS ship attribute data set are removed, and the night vision target feature data with a PSNR greater than 35dB are defogged and motion blur compensated to obtain the multi-source standardized preprocessing data of ships.

[0017] This invention deploys X-band navigation radars on the vessel's deck and mast to collect data such as target position, distance, speed, and projected area. The radar's raw point cloud data provides detailed target detection information, particularly at longer ranges. This enhances the vessel's long-range detection capabilities in complex waters and accurately identifies the target's dynamic characteristics, such as position and speed. This provides important spatial data support for subsequent environmental perception and decision-making. Furthermore, by collecting the vessel's MMSI number, speed, heading, and static information through an AIS receiver, surrounding vessels can be accurately identified and their behavior predicted based on their dynamic tracks, enhancing navigation safety. Furthermore, the installation of night vision devices provides target detection capabilities in low-light environments. Specifically, by extracting target outline, type, and relative distance data, this effectively addresses the challenges of navigation at night or in adverse weather conditions, ensuring accurate target positioning even in suboptimal visual conditions. This information, coupled with real-time data provided by the Beidou navigation system, further enhances the vessel's spatial positioning and attitude information, enabling the vessel to continuously monitor its status and ensure accurate and stable navigation. Through standardized preprocessing of this multi-source data, including sea clutter suppression for radar point cloud data, timeout signal removal for AIS data, and defogging and motion blur compensation for night vision images, the high quality and reliability of the input data are ensured, noise interference is reduced, and data validity is improved, thereby providing clear and accurate input data for the vessel's high-precision decision-making system. Overall, the technical means of these steps have significantly improved the vessel's ability to perceive its surroundings, while ensuring efficient data processing and accurate information fusion, thereby enhancing the vessel's navigation safety and collaborative combat capabilities in complex and dynamic environments.

[0018] Preferably, step S2 includes the following steps:

[0019] Step S21: mapping the multi-source standardized pre-processed data of the vessel into a three-dimensional space and constructing a target three-dimensional model;

[0020] Step S22: Outputting target semantic label data using the target three-dimensional model;

[0021] Step S23: The target semantic label data is fused with the Beidou navigation status data set, and the vessel's own motion drift is performed through a Kalman filter with a prediction period of 100ms to obtain an environmental perception matrix.

[0022] By mapping the multi-source standardized pre-processed data of a vessel into three-dimensional space and constructing a three-dimensional target model, this method converts data from various sensors (such as radar, AIS, and night vision devices) into a unified three-dimensional coordinate representation. This step assigns a specific spatial location to each target, achieving high-precision modeling of the surrounding environment, ensuring accurate target detection and clear representation of spatial relationships. Secondly, the output of the three-dimensional target model generates target semantic label data, which enables the vessel system to classify and identify various targets in the environment. The target semantic label not only includes the target's physical attributes (such as position and velocity) but also its category (such as ship, obstacle, etc.), facilitating intelligent decision-making and path planning for the vessel. Furthermore, the target semantic label data is integrated with the Beidou navigation status dataset, enabling the vessel's position, heading, and velocity information to work in conjunction with the surrounding target data, enhancing the system's global perception of the vessel's position and target status. The application of a Kalman filter predicts and corrects the vessel's own motion drift, providing a short-term prediction cycle of 100ms, ensuring real-time updates and accuracy of the environmental perception matrix. This predictive data fusion method effectively reduces errors introduced during motion and improves the accuracy and stability of navigation control, especially in complex environments or rapidly changing conditions. Furthermore, the generation of an environmental perception matrix provides comprehensive spatial and dynamic information support for subsequent navigation decisions and obstacle avoidance systems, making it a key component of intelligent vessel decision-making. Therefore, the technical means of these steps demonstrate high efficiency and accuracy in data processing and fusion, significantly enhancing the vessel's environmental perception and adaptive adjustment capabilities, thereby improving navigation safety and decision-making efficiency.

[0023] Preferably, step S21 includes the following steps:

[0024] Step S211: Convert the polar coordinates of the radar point cloud data into the local radar ENU coordinate system, and obtain WGS-84 radar geographic coordinates through coordinate rotation and translation; perform WGS-84 coordinate system conversion based on the AIS ship attribute dataset to obtain WGS-84AIS coordinates; extract camera calibration parameters based on the night vision target feature data, and back-project the pixel coordinates into the local 3D coordinate system for local WGS-84 conversion to obtain WGS-84 night vision device coordinates;

[0025] Step S212: Distance weights are assigned according to the weights of WGS-84 radar geographic coordinates being 0.6, WGS-84 AIS coordinates being 0.3, and WGS-84 night vision device coordinates being 0.1, and the data are mapped to three-dimensional space to obtain vessel three-dimensional space coordinate fusion data;

[0026] Step S213: Screen the fused data based on the three-dimensional spatial coordinate fusion data of the vessel with a position error less than 2m, and construct the model using a preset gradient loss function to obtain a target three-dimensional model.

[0027] The present invention converts radar point cloud data from polar coordinates to a local radar ENU coordinate system, and further derives WGS-84 radar geographic coordinates through coordinate rotation and translation. This allows the radar data to be accurately aligned with the global standard coordinate system, thus providing high-precision spatial information for vessel positioning. The WGS-84 coordinate conversion of the AIS vessel attribute dataset provides another data source to supplement the dynamically changing position of the vessel during navigation and ensures consistency with the radar data. Furthermore, by performing camera calibration on night vision target feature data and back-projecting its pixel coordinates to a local three-dimensional coordinate system for WGS-84 coordinate conversion, the vessel's target recognition and position calibration capabilities in low-light or nighttime environments are further enhanced. Through these conversion steps, data generated by various sensors (radar, AIS, and night vision devices) can be unified in a standard coordinate system for fusion and application. Secondly, step S212 assigns weights to different data sources (radar, AIS, and night vision devices) for three-dimensional spatial mapping, enabling the vessel's positioning system to comprehensively consider the reliability and contribution of each data source, further optimizing the accuracy of spatial coordinates and the fusion effect. This weighted fusion method enables the system to effectively coordinate and optimize between multiple sources of information, reducing the deviation introduced by the instability or error of a single sensor. Finally, step S213 screens for positions with an error of less than 2m and uses a preset gradient loss function to build the model, ensuring that the generated target three-dimensional model has high precision and meets the needs of actual applications. In this process, by optimizing the relationship between the target model and the data accuracy, the spatial perception and decision-making capabilities of the vessel for the environment are further improved, thereby achieving more reliable navigation control and safety assurance. Therefore, these technical means provide the vessel with a high-precision, real-time updated three-dimensional environmental perception model through sophisticated data conversion, weighted fusion and error screening, significantly improving its adaptability and reliability in complex sea conditions and dynamic environments.

[0028] Preferably, step S3 of channel coding the environment sensing matrix and performing subcarrier allocation includes the following:

[0029] Inputting the environment perception matrix into the first convolutional encoder to generate a coding polynomial to obtain the environment perception first convolutional code;

[0030] Using an interleaver to block-interleave the first environmental-aware convolutional code to obtain environmental-aware interleaver data;

[0031] Inputting the environment-aware interleaver data into the second encoder for convolution to obtain an environment-aware parallel convolutional code;

[0032] Subcarrier allocation is performed on the environment-aware parallel convolutional code to obtain a multi-carrier modulated signal, wherein the subcarrier allocation on the environment-aware parallel convolutional code includes the following steps:

[0033] The environment-aware parallel convolutional code adopts 64QAM SNR subcarrier modulation to obtain the environment 64-bit SNR subcarrier signal data; the environment-aware parallel convolutional code adopts 16QAM SNR subcarrier modulation to obtain the environment 16-bit subcarrier signal data; the environment-aware parallel convolutional code adopts QPSK SNR subcarrier modulation to obtain the environment QPSK subcarrier signal data.

[0034] The present invention uses a first-channel convolutional encoder to encode the environmental perception matrix and generate a polynomial convolutional code. This process converts the original data into a highly redundant form through convolutional coding, effectively enhancing the robustness of the signal during transmission. This redundancy allows the receiver to recover the original data through error correction algorithms in the event of signal interference or loss, thereby improving system reliability. Next, a 1024-bit interleaver performs block interleaving on the convolutional code, completing the data interleaving process. Interleaving mitigates burst errors during data transmission. By rearranging adjacent bits, it reduces the impact of consecutive transmission errors on the data block. This operation is particularly suitable for environments with high bit error rates, further improving transmission stability and accuracy. The interleaved environmental perception data is then input into a second-channel encoder for further convolution processing, generating a parallel convolutional code. This process further enhances data redundancy, improves noise and interference resistance, and lays a solid foundation for subsequent modulation and transmission. Subcarrier allocation is then performed, and the parallel convolutional code is modulated according to different signal-to-noise ratio (SNR) conditions. During this process, high-SNR subcarriers utilize 64QAM modulation, enabling higher data transmission rates in high SNR environments. Medium-SNR subcarriers utilize 16QAM modulation for medium SNR environments, and low-SNR subcarriers utilize QPSK modulation to ensure reliable communication even in low SNR environments. This multi-layered, multi-SNR modulation scheme allows the system to dynamically adjust signal transmission methods based on the actual transmission environment, thereby optimizing signal quality and transmission efficiency under varying channel conditions. Ultimately, these steps optimize signal transmission, enabling more stable and efficient data transmission in complex environments, effectively reducing signal interference and error rates, and ensuring communication reliability and data integrity. These technical measures not only ensure data transmission stability but also enhance the system's adaptability to varying channel conditions, possessing significant practical application value.

[0035] Preferably, step S3 of encapsulating IP data packets according to the multi-carrier modulated signal and relaying the IP data packets by the wireless broadband ad hoc network radio station includes the following:

[0036] The multi-carrier modulation signal is divided into time slots according to 10ms / time slot, and priority queue scheduling is performed to obtain priority queue management data, where the priority queue management data includes situation data, voice data and basic ship parameter data;

[0037] Preemptive access is performed on situation data, and time slot resources are directly allocated. Binary exponential backoff is performed on voice data to allocate time slot resources. The remaining time slot resources are competed for and transmitted for basic vessel parameter data.

[0038] Encapsulate IP data packets according to the priority queue management data, and append the target node ID and hop limit fields to the packet header to obtain multi-carrier encapsulated data of the vessel;

[0039] The wireless broadband ad hoc network radio is used to forward the multi-carrier encapsulated data of the ships in multiple channels according to the working frequency band of 2.4 GHz, bandwidth of 32 MHz and transmission filter of 20W. The relay nodes are selected according to the target node ID and hop limit to obtain the shared data of the ship formation situation.

[0040] The present invention effectively improves the data transmission efficiency and real-time performance of multi-carrier communication systems in complex environments through refined time slot division and priority queue scheduling. First, according to the 10ms time slot division, the system can achieve precise control of the data flow, and through priority queue scheduling, different types of data are divided into situation data, voice data, and basic vessel parameter data. Preemptive access to situation data ensures that high-priority information can quickly occupy time slot resources and ensure that situation information can be transmitted in real time, thereby improving the system's responsiveness in emergency situations. Voice data is allocated time slot resources through a binary exponential backoff mechanism. This mechanism can flexibly adjust time slot allocation when multiple users are competing, avoiding conflicts and delays in voice data and ensuring smooth voice communication. Basic vessel parameter data uses the remaining time slot resources for competitive transmission. This strategy effectively utilizes idle time slots and ensures the transmission efficiency of basic vessel parameter data without affecting high-priority data transmission. This multi-level, priority scheduling method ensures the effective transmission of various data streams and dynamically adjusts transmission resources according to actual needs, optimizing system performance. Next, the system encapsulates the priority queue management data into IP packets and appends the destination node ID and hop limit fields to the packet header. This process enhances data routing, enabling data to be forwarded across multiple wireless broadband ad hoc network radios. In this process, utilizing the 2.4 GHz operating frequency band and 32 MHz bandwidth, combined with 20 W of transmit power, the system enables data transmission over a wide range. Relay nodes are selected based on the destination node ID and hop limit, ensuring that data is efficiently and accurately transmitted to the target node. This mechanism not only optimizes the signal propagation path but also ensures stable data transmission in complex environments. In summary, these technical steps significantly improve the data transmission efficiency and reliability of the vessel formation system through precise resource allocation, dynamic time slot management, and efficient data encapsulation and forwarding strategies. This helps ensure the real-time and accuracy of shared vessel formation situation data, thereby enhancing the formation's collaborative combat capabilities.

[0041] Preferably, step S4 includes the following steps:

[0042] Step S41: Generate a global initial route for the shared data of the boat formation situation according to the voyage cost coefficient of 0.7 and the safety distance cost coefficient of 0.3;

[0043] Step S42: Perform local resolution screening on the global initial route to obtain a local obstacle avoidance path, wherein the evaluation indicators for local resolution screening are collision avoidance rule weight 0.5, navigation constraint weight 0.3, environmental interference weight 0.2, speed resolution 0.1 m / s, and heading angle resolution 1°;

[0044] Step S43: Marking route nodes based on the local obstacle avoidance path and generating a formation cooperative navigation instruction set.

[0045] The present invention generates a global initial route for shared data on the status of a fleet of vessels by weighting the range cost coefficient (0.7) and the safety distance cost coefficient (0.3), ensuring the rationality and feasibility of route planning. The range cost coefficient primarily considers the optimization of the range between vessels, while the safety distance cost coefficient ensures a safe distance between vessels, avoiding the risk of collision during the formation's navigation. This strategy, through the rational configuration of weights at the data level, makes the overall navigation path of the fleet both economical and safe. The global initial route is screened for local resolution using a preset distributed obstacle avoidance comprehensive evaluation function, further optimizing the safety of the navigation path. This evaluation function evaluates the path based on the collision avoidance rule weight (0.5), navigation constraint weight (0.3), and environmental interference weight (0.2). By setting these weights at the data level, a reasonable balance is achieved between collision avoidance requirements, navigation constraints, and the impact of environmental interference. During local resolution screening, the speed resolution (0.1 m / s) and heading angle resolution (1°) are set to achieve more precise and accurate route planning, enabling real-time response to various obstacles and changes encountered by vessels in dynamic environments. Through this meticulous screening, the system generates a local obstacle avoidance path with effective performance, ensuring safe and efficient navigation of the fleet in complex environments. Based on the local obstacle avoidance path, key route nodes are marked, and a coordinated navigation instruction set for the formation is generated. This process precisely identifies key route nodes, enabling the formation to achieve precise navigation adjustments and efficient collaborative operations when performing coordinated missions. From a data perspective, the marking of key routes ensures that the formation can dynamically adjust to the actual route conditions during mission execution to address changes or obstacles during navigation. The generation of instruction sets enables coordinated operations between vessels, thereby optimizing navigation efficiency and safety. Overall, this series of steps, through efficient data processing and precise control, ensures safe and coordinated navigation of the fleet in complex environments, thereby improving the combat effectiveness of the entire formation.

[0046] Preferably, step S5 includes the following steps:

[0047] Step S51: Overlaying and displaying nautical charts according to the formation cooperative navigation instruction set to obtain a real-time route situation map;

[0048] Step S52: Performing a yaw-collision avoidance warning based on the real-time route situation map to obtain a level 3 warning signal;

[0049] Step S53: Based on the third-level warning signal and the real-time route situation map, control instructions are issued at a baud rate of 1 Mbps and a frame interval of 10 μs to obtain a normal control instruction set; if the packet loss rate of the ad hoc network is greater than 5%, it is determined that the communication link is interrupted and a degraded control instruction set is generated;

[0050] Step S54: constructing a vessel wireless communication control report based on the normal control instruction set and the degraded control instruction set.

[0051] This invention significantly improves the navigation safety and control reliability of fleets in complex environments through dynamic situational awareness, real-time early warning, communication anomaly handling, and command feedback mechanisms. First, a real-time route situation map is generated by overlaying nautical charts based on the fleet's collaborative navigation command set. This allows vessels to dynamically monitor their navigation status in an intuitive data visualization environment, enhancing their overall understanding of the navigation environment and command execution. Based on the real-time route situation map, intelligent discrimination of yaw and collision avoidance risks is performed. A three-level early warning signal system is established through multidimensional data analysis, thereby refining the navigation risk level at the data level and improving the accuracy and timeliness of early warning information processing. Next, in the communication control link, a high-speed command issuance mechanism with a baud rate of 1 Mbps and a frame interval of 10 μs enables high-frequency, low-latency command transmission, ensuring the efficiency and stability of the command link under normal conditions. Furthermore, a monitoring threshold for an ad hoc network packet loss rate exceeding 5% is designed. If this threshold is triggered, a communication link interruption is quickly detected and a downgraded control command set is generated. This ensures minimum control capabilities during link instability, demonstrating the system's intelligent response to communication anomalies and fault-tolerant handling capabilities. Finally, the normal and degraded control instruction sets are consolidated to generate a vessel wireless communication control report, enabling systematic recording and backtracking of control instruction integrity, improving data traceability and fault tracing capabilities for navigation management. Overall, this process efficiently integrates navigation status data, warning information, communication link status, and control instruction flows to build a closed-loop, data-driven control system. This not only enhances the fleet's navigation autonomy and risk mitigation capabilities, but also provides a solid data foundation for subsequent decision-making optimization and intelligent navigation.

[0052] Preferably, step S52 includes the following:

[0053] Based on the real-time route situation map, the ship can make a yaw judgment and perform CPA calculation to obtain the ship's CPA data; based on the real-time route situation map, the ship can make a collision avoidance prediction and perform TCPA calculation to obtain the ship's TCPA data;

[0054] When the CPA data is greater than 0.5 nautical miles or the TCPA data is greater than 300s, it is judged as a green normal signal;

[0055] When the CPA data is between 0.3 nautical miles and 0.5 nautical miles and the TCPA data is between 100s and 300s, it is judged as a yellow caution signal;

[0056] When the CPA data is less than 0.3 nautical miles or the TCPA data is less than 300s, it is judged as an emergency red signal.

[0057] This invention constructs a quantitatively-based yaw and collision avoidance warning system by extracting and calculating CPA (Closest Point Approach) and TCPA (Closest Point Approach Time) data in real time, enabling efficient identification and categorized response to navigation conflict risks. First, the system dynamically extracts the CPA and TCPA values ​​between vessels and potential conflict targets from a real-time route situation map, forming a continuous data stream for navigation safety assessment, providing accurate and timely fundamental data support for subsequent decision-making. Subsequently, a tiered assessment criteria is established: CPA values ​​greater than 0.5 nautical miles or TCPA values ​​greater than 300 seconds are labeled as green, indicating sufficient safety separation between vessels and clearly defining the safety baseline for navigation status. When CPA values ​​are between 0.3 and 0.5 nautical miles and TCPA values ​​between 100 and 300 seconds, the system triggers a yellow caution signal, indicating a moderate risk requiring attention from the pilot or automated decision-making system. If CPA values ​​are less than 0.3 nautical miles or TCPA values ​​are less than 100 seconds, a red emergency signal is triggered, providing rapid data-based warning of potential high-risk collisions. This method of hierarchical processing based on specific numerical thresholds not only enhances the system's ability to sensitively capture dynamic changes in navigation risks, but also makes risk warnings highly interpretable and real-time, avoiding safety hazards caused by data ambiguity or delayed responses. In addition, clear grading standards also make it easier to use warning signals as input to drive subsequent obstacle avoidance decisions, path adjustments, or control instructions, further forming a data-driven closed-loop safety control mechanism. Overall, this process significantly improves the intelligence and dataization level of ship navigation safety management through the precise calculation, scientific grading, and real-time application of CPA and TCPA data, providing reliable support for efficient collision avoidance and risk control in formation navigation environments. BRIEF DESCRIPTION OF THE DRAWINGS

[0058] Figure 1 A schematic flow chart of a method for controlling intelligent wireless communication of a vessel is provided;

[0059] Figure 2 for Figure 1 Detailed implementation steps of step S5;

[0060] The purpose, features and advantages of the present invention will be further described with reference to the accompanying drawings and in conjunction with the embodiments. DETAILED DESCRIPTION

[0061] The following is a clear and complete description of the technical method of the present invention in conjunction with the accompanying drawings. Obviously, the embodiments described are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without making any creative work are within the scope of protection of the present invention.

[0062] In addition, the accompanying drawings are merely schematic illustrations of the present invention and are not necessarily drawn to scale. Identical reference numerals in the figures denote identical or similar parts, and thus repetitive descriptions thereof will be omitted. Some of the block diagrams shown in the accompanying drawings are functional entities that do not necessarily correspond to physically or logically separate entities. These functional entities may be implemented in software, in one or more hardware modules or integrated circuits, or in different network and / or processor and / or microcontroller approaches.

[0063] It should be understood that although the terms "first," "second," and the like may be used herein to describe various elements, these elements should not be limited by these terms. These terms are used solely to distinguish one element from another. For example, a first element may be referred to as a second element, and similarly, a second element may be referred to as a first element, without departing from the scope of the exemplary embodiments. The term "and / or" as used herein includes any and all combinations of one or more of the listed associated items.

[0064] To achieve this, please refer to Figures 1 to 2 , a method for controlling intelligent wireless communication for a vessel, the method comprising the following steps:

[0065] Step S1: Collect radar raw point cloud data, AIS ship attribute data set, Beidou navigation status data set and night vision target feature data through multi-dimensional sensors to construct multi-source standardized pre-processed data of ships;

[0066] Step S2: Perform three-dimensional spatial mapping on the multi-source standardized pre-processed data of the vessel and construct a target three-dimensional model; and construct an environment perception matrix based on the target three-dimensional model;

[0067] Step S3: Channel coding is performed on the environment perception matrix, and subcarrier allocation is performed to obtain a multi-carrier modulated signal; IP data packets are encapsulated according to the multi-carrier modulated signal, and relayed and transmitted by a wireless broadband ad hoc network radio station to obtain shared data on the situation of the ship formation;

[0068] Step S4: Generate a global initial route based on the shared data of the boat formation situation and construct a local obstacle avoidance path; mark the route nodes according to the local obstacle avoidance path and generate a formation collaborative navigation instruction set;

[0069] Step S5: The nautical chart is superimposed and displayed according to the formation collaborative navigation instruction set to obtain a real-time route situation map; a deviation-collision avoidance warning is performed on the real-time route situation map to obtain a three-level warning signal; a control instruction is issued to the central control system based on the three-level warning signal and the real-time route situation map to obtain a vessel wireless communication control report.

[0070] In the embodiment of the present invention, reference Figure 1 FIG. 1 is a flow chart of a method for controlling a boat's intelligent wireless communication according to the present invention. In this example, the method for controlling a boat's intelligent wireless communication includes the following steps:

[0071] Step S1: Collect radar raw point cloud data, AIS ship attribute data set, Beidou navigation status data set and night vision target feature data through multi-dimensional sensors to construct multi-source standardized pre-processed data of ships;

[0072] In this embodiment of the present invention, a multi-dimensional sensor array is used to collect raw radar point cloud data of the vessel's surrounding environment, vessel attribute data from the AIS (Automatic Identification System), real-time positioning and heading status data from the Beidou navigation system, and target feature information obtained by the night vision imaging system. The radar raw point cloud data is collected using a pulse ranging method. By transmitting high-frequency electromagnetic waves and receiving reflected signals, the time delay and intensity of each echo are recorded to form a dense set of spatially discrete points. In its raw state, the point cloud data contains basic features such as distance, reflection intensity, azimuth, and pitch angle. The AIS dataset is collected by the Automatic Identification System for Ships and includes, but is not limited to, the vessel's unique identification code, vessel type, speed, heading, draft, and dynamic status. This data is received and stored in real time using a standardized message structure (such as the ITU-R M.1371 standard). The Beidou navigation status dataset, based on satellite positioning system output, contains high-precision latitude and longitude coordinates, ground speed vectors, track angles, attitude information, and timestamps. The data format complies with NMEA-0183 or a custom high-precision protocol, and uses differential or RTK corrections to enhance spatial positioning accuracy. Night vision target feature data is collected using short-wave infrared or long-wave infrared imaging devices. Preliminary target region segmentation is performed based on features such as image intensity histograms, target contour gradient variations, and thermal image partition contrast. The dataset then outputs a pixel-level grayscale matrix and contour annotation information. To ensure consistency and fusion of multi-source data, standardized preprocessing is required for each data source. This includes time synchronization (using a unified clock source or high-precision timestamp correction), spatial registration (using sensor extrinsic calibration and coordinate transformation matrices to unify data to a geographic coordinate system or inertial reference frame), data resampling (using uniform interpolation or filtering compensation based on a specified time window), and feature normalization (dimensionalization for different data scales). This ultimately generates standardized multi-source preprocessed data for the vessel, providing a unified and standardized data input source for subsequent 3D spatial mapping and environmental modeling.

[0073] Step S2: Perform three-dimensional spatial mapping on the multi-source standardized pre-processed data of the vessel and construct a target three-dimensional model; and construct an environment perception matrix based on the target three-dimensional model;

[0074] In an embodiment of the present invention, based on standardized multi-source pre-processed data, the vessel and its surrounding environment are first reconstructed using a three-dimensional spatial mapping method. Specifically, the radar point cloud data is projected into a three-dimensional position using a spherical coordinate to rectangular coordinate formula based on the original polar coordinate information. Static dimension parameters (such as length, width, and height) and real-time dynamic parameters (such as speed and heading) in the AIS ship attribute data are used to label the point cloud data and fit the target voxel boundaries. Beidou navigation status data provides a precise global positioning reference for the point cloud collection at each moment, ensuring the geographic consistency and dynamic continuity of the three-dimensional mapping results. Night vision target feature data undergoes feature correlation analysis between the image space and the point cloud space, and through nearest neighbor search and feature descriptor matching, it further enhances the compensated recognition of targets in weak echo or occluded areas. After the three-dimensional spatial mapping is completed, a high-dimensional environmental perception matrix is ​​constructed based on the spatial position, morphological characteristics, motion state and surface reflection characteristics of each independent target voxel. The matrix is ​​usually stored in a sparse tensor or compressed adjacency list in terms of data structure. The internal elements represent the target existence probability, category confidence, motion trend vector and local environmental feature coding of each spatial unit, providing efficient data support for subsequent communication signal processing and route planning.

[0075] Step S3: Channel coding is performed on the environment perception matrix, and subcarrier allocation is performed to obtain a multi-carrier modulated signal; IP data packets are encapsulated according to the multi-carrier modulated signal, and relayed by a wireless broadband ad hoc network radio station to obtain shared data on the situation of the ship formation;

[0076] In this embodiment of the present invention, adaptive channel coding is first performed on the constructed environmental perception matrix. The specific process involves extracting the priority and update frequency indicators of each target unit in the perception matrix and selecting an appropriate channel coding scheme, such as LDPC (Low-Density Parity Check) or Turbo coding, based on dynamic channel conditions (such as signal-to-noise ratio, interference index, and bandwidth resource allocation), to achieve a balanced optimization between error robustness and transmission efficiency. After channel coding, the coded perception matrix is ​​divided into several logical signal blocks according to frequency domain resource mapping rules and assigned to subcarriers in a multicarrier system. Multicarrier modulation is performed using OFDM (Orthogonal Frequency Division Multiplexing) to form a parallel data transmission structure. The modulated signal is then encapsulated according to the IP datagram protocol standard to construct a data frame structure, including frame header definition, payload packaging, and error detection code addition. The encapsulated IP data packets are relayed via wireless broadband ad hoc network radios. Leveraging link state awareness and dynamic routing reconfiguration mechanisms, they ensure low-latency and highly reliable transmission of shared data on formation status between vessels. During the entire process, each node needs to update the routing table and channel status table in real time to support adaptive data relay in a dynamic environment.

[0077] Step S4: Generate a global initial route based on the shared data of the boat formation situation and construct a local obstacle avoidance path; mark the route nodes according to the local obstacle avoidance path and generate a formation collaborative navigation instruction set;

[0078] In this embodiment of the present invention, based on the acquired shared data on the vessel formation situation, an initial route is first generated using a global path search algorithm. This process uses each vessel's current position, target position, and static obstacle information from the environmental perception matrix as input. A heuristic search algorithm (such as the A algorithm or the DLite algorithm) is used to search for the optimal connected path in three-dimensional space. The path is then adjusted for feasibility based on the vessel's navigation characteristics (such as minimum turning radius, maximum acceleration, and navigation inertia). After the initial route is generated, a local obstacle avoidance path planning module is activated to address dynamically changing components of the environmental perception matrix (such as moving targets and sudden obstacles). This module uses local feasible region reconstruction and rolling window path optimization (e.g., based on the DWA dynamic window method) to adjust the current route nodes in real time. Subsequently, after the global and local paths are integrated, important nodes on the path are marked based on key turning points, speed change nodes, and coordinated formation requirements. Node attributes such as position coordinates, speed change instructions, and obstacle avoidance priority are recorded. Finally, by integrating the information of each node and the formation navigation rules, a standardized formation collaborative navigation instruction set is generated. The instruction set is organized in a hierarchical structure, including single-boat autonomous navigation instructions, formation synchronization adjustment instructions, and formation obstacle avoidance coordination instructions, providing an instruction input source for subsequent track display and early warning control.

[0079] Step S5: The nautical chart is superimposed and displayed according to the formation collaborative navigation instruction set to obtain a real-time route situation map; a deviation-collision avoidance warning is performed on the real-time route situation map to obtain a three-level warning signal; a control instruction is issued to the central control system based on the three-level warning signal and the real-time route situation map to obtain a vessel wireless communication control report.

[0080] In this embodiment of the present invention, based on the generated formation cooperative navigation instruction set, the route data is first fused and overlaid with the digital nautical chart. Specifically, by parsing the important node sequences in the navigation instruction set, their coordinate information, course change instructions, and speed adjustment information are mapped to the electronic nautical chart data layer. Spatial alignment and projection unification are performed using the geographic reference framework in the standard S-57 or S-101 electronic nautical chart data format, forming a real-time route situation map. The situation map is organized as a vector layer stack, with track lines, node markers, and obstacle avoidance path segments managed as independent layers. Metadata tags such as speed, course, and node attributes are also attached to ensure the integrity and searchability of the track data. After the situation map is generated, the deviation and collision avoidance warning module is activated to perform real-time comparison and analysis of the current real-time route with the dynamic target trajectory in the environmental perception matrix. The potential collision risk level is calculated using a multi-level risk assessment model (e.g., based on the CPA closest point distance and TCPA closest time arrival model). According to the evaluation results, there are three levels of warning signals, namely primary warning, intermediate warning and advanced warning. Each level of signal is defined by different time windows, spatial thresholds and track deviations, and a warning notification is generated in the form of a data structured message. While generating the warning signal, a control instruction set is dynamically generated based on the real-time situation map and the warning level, including adjusting the heading, slowing down, changing the route or issuing collaborative obstacle avoidance instructions. The control instructions are packaged in a standard communication protocol and sent to each vessel terminal through a wireless communication link. Finally, all warning signals and corresponding control measures are recorded and organized to generate a vessel wireless communication control report. The report lists in detail the warning trigger time, trigger position, corresponding control response instructions and execution feedback status, which serves as the basic data source for situation monitoring and navigation safety audits.

[0081] Preferably, step S1 includes the following steps:

[0082] Step S11: deploying navigation radars on the deck and mast of the vessel, operating in the X-band frequency band, with a detection range greater than 20 nautical miles and a resolution less than 30 meters, to scan the surrounding sea area, collecting target azimuth, distance, speed, and projected area data to obtain radar raw point cloud data;

[0083] Step S12: Using a one-way AIS receiver with a frequency of 161.975 MHz and a decoding rate of 1 second, the MMSI number, speed, heading, and static information of the ship are collected to generate an AIS ship attribute data set.

[0084] Step S13: Installing a visible light + infrared dual-band night vision device on the front of the boat to collect optical images in night or low-light environments, and extracting target outline, type and relative distance data to obtain night vision target feature data;

[0085] Step S14: Using the BeiDou-3 navigation terminal, the vessel's latitude and longitude, speed, and attitude angle data are collected in real time to generate a BeiDou navigation status data set;

[0086] Step S15: Check the radar point cloud data for a false alarm rate less than or equal to The sea clutter is suppressed, the timeout and unupdated signals of the AIS ship attribute data set are removed, and the night vision target feature data with a PSNR greater than 35dB are defogged and motion blur compensated to obtain the multi-source standardized preprocessing data of ships.

[0087] In an embodiment of the present invention, a navigation radar with an X-band operating frequency band is deployed on the deck and mast of a ship to detect sea areas within a range of more than 20 nautical miles. The radar system synchronously records the target's azimuth, distance, radial velocity and projected area information in each pulse repetition period, and forms an original point cloud data set through the amplitude, frequency offset and time delay of the echo signal; then, in step S12, a one-way automatic identification system (AIS) receiver with an operating frequency of 161.975 MHz and a decoding rate of once per second is used to receive the MMSI number, current speed, heading information, and static attributes such as ship type and size broadcast by surrounding ships in real time, and signal unpacking, error detection and field analysis are performed at the receiving end to generate an AIS ship attribute data set; in step S13, a visible light and infrared dual-band radar is installed at the front end of the ship. The band-fused night vision imaging equipment collects continuous image sequences in low-light or nighttime environments in the optical and thermal infrared channels respectively, extracts the target contour boundary through the image segmentation algorithm, and infers the target type and its relative distance by combining the dual-band feature matching and stereo disparity estimation technology to obtain the night vision target feature data; at the same time, in step S14, by installing the BeiDou-3 high-precision navigation terminal, the latitude and longitude, speed and attitude angle (including roll angle, pitch angle, yaw angle) of the ship are collected in real time. After differential correction and inertial measurement unit (IMU) auxiliary fusion, a high-update frequency and low-latency BeiDou navigation status data set is generated; finally, in step S15, each source data is standardized. Among them, the radar point cloud data adopts the sea clutter suppression algorithm based on the constant false alarm rate (CFAR) detector to ensure that the false alarm rate is controlled at 10⁻ 6Next, the detected point cloud echoes are simultaneously screened for a second time based on signal strength and velocity thresholds. AIS ship attribute datasets undergo consistency checks based on reception timestamps, removing unupdated signals that exceed a preset timeout threshold to avoid interference from historical data. After screening image frames with PSNR greater than 35dB for night vision target feature data, a dark channel prior (DCP)-based method is used for image dehazing. This is combined with a motion blur compensation algorithm based on blur kernel estimation and inverse convolution to correct for dynamic blur distortion in the imaging. This completes standardized preprocessing of radar, AIS, night vision, and navigation multi-source data in terms of data consistency, temporal synchronization, and feature integrity, laying the foundation for subsequent data fusion and target recognition.

[0088] Preferably, step S2 includes the following steps:

[0089] Step S21: mapping the multi-source standardized pre-processed data of the vessel into a three-dimensional space and constructing a target three-dimensional model;

[0090] Step S22: Outputting target semantic label data using the target three-dimensional model;

[0091] Step S23: The target semantic label data is fused with the Beidou navigation status data set, and the vessel's own motion drift is performed through a Kalman filter with a prediction period of 100ms to obtain an environmental perception matrix.

[0092] In one embodiment of the present invention, three-dimensional spatial mapping is performed based on standardized pre-processed multi-source vessel data. Specifically, this involves transforming the radar point cloud data into a geographic coordinate system (e.g., WGS-84) and performing attitude correction. The pitch, roll, and yaw errors of the point cloud are corrected using attitude angle data, achieving spatial alignment between the radar observation data and the navigation data. Simultaneously, target contours and relative distance information extracted from the night vision target feature data are also geo-backprojected based on real-time navigation position information and unified into the same three-dimensional coordinate system, thereby constructing a real-world three-dimensional model of the target. The three-dimensional model is organized in the form of a sparse point cloud or a dense grid, with nodes containing multi-dimensional attributes such as spatial position, velocity vector, and target type. Subsequently, in step S22, target semantic labels are generated for the constructed three-dimensional model using a method based on deep feature extraction and clustering. Specifically, a classifier based on features such as point cloud density, geometry, and motion state is used to classify the targets into categories such as ships, buoys, and sea debris. Attribute information such as target ID, category label, size estimate, and velocity direction is simultaneously output, forming a structured target semantic label dataset. Next, in step S23, the target semantic label data is fused with the Beidou navigation status dataset. During the fusion process, asynchronous data sampling issues are addressed through timestamp alignment and position interpolation methods to ensure consistency between track and target data under the same time reference. To further correct position deviations caused by navigation system drift, sensor errors, or motion disturbances, a Kalman filter with a prediction period of 100ms is used to estimate the vessel's own motion state in real time. The filter uses the current navigation state as a priori estimate and sensor fusion data as observation updates. Through the joint optimization of the state transition matrix and the observation matrix, track drift is corrected and target trajectory changes are smoothed within each prediction period. The final output is an environmental perception matrix that integrates position, velocity, attitude, and target classification information. This matrix is ​​organized as a time series structure to support the data requirements of subsequent route planning and dynamic situation deduction.

[0093] Preferably, step S21 includes the following steps:

[0094] Step S211: Convert the polar coordinates of the radar point cloud data into the local radar ENU coordinate system, and obtain WGS-84 radar geographic coordinates through coordinate rotation and translation; perform WGS-84 coordinate system conversion based on the AIS ship attribute dataset to obtain WGS-84AIS coordinates; extract camera calibration parameters based on the night vision target feature data, and back-project the pixel coordinates into the local 3D coordinate system for local WGS-84 conversion to obtain WGS-84 night vision device coordinates;

[0095] Step S212: Distance weights are assigned according to the weights of WGS-84 radar geographic coordinates being 0.6, WGS-84 AIS coordinates being 0.3, and WGS-84 night vision device coordinates being 0.1, and the data are mapped to three-dimensional space to obtain vessel three-dimensional space coordinate fusion data;

[0096] Step S213: Screen the fused data based on the three-dimensional spatial coordinate fusion data of the vessel with a position error less than 2m, and construct the model using a preset gradient loss function to obtain a target three-dimensional model.

[0097] In an embodiment of the present invention, the radar raw point cloud data is extracted from its distance, azimuth, and elevation information in the polar coordinate system. This information is then converted to the local East-North-Sky (ENU) coordinate system using a spherical-to-Rectangular coordinate transformation formula. A coordinate rotation matrix and displacement vector are then applied to the vessel attitude angle data to complete the coordinate projection from the local radar coordinate system to the global WGS-84 geographic coordinate system, thereby obtaining WGS-84 radar geographic coordinates. Simultaneously, the latitude and longitude information of each target in the AIS vessel attribute data is directly converted based on the parameters of the Earth ellipsoid model to form WGS-84 AIS coordinates that are standardized with the radar data. For night vision target feature data, camera calibration parameters are extracted, including an internal parameter matrix (focal length, principal point, distortion coefficient) and an external parameter matrix (rotation matrix and translation vector). A back-projection algorithm is then used to restore the night vision device pixel coordinates to the local three-dimensional coordinate system. After attitude correction and georeferencing, the WGS-84 night vision device coordinates are obtained. In step S212, radar coordinates are weighted 0.6, AIS coordinates 0.3, and night vision device coordinates 0.1, respectively, for the three different data sources. Spatial mapping is performed using the weighted Euclidean distance method, completing multi-source spatial fusion processing to form a unified 3D spatial coordinate fusion dataset describing the vessel and surrounding targets. During the spatial fusion process, nearest neighbor matching and coordinate weight superposition are performed using synchronized timestamps and a spatial neighborhood search algorithm (such as a kd-tree) to ensure consistency of multi-source observations and optimize local accuracy. Next, in step S213, a position error screening mechanism is applied to the 3D spatial coordinate fusion dataset. Fusion nodes with excessive errors are removed based on a pre-set position error threshold (less than 2 meters) to improve subsequent model accuracy. Furthermore, a gradient loss function is defined using the spatial continuity and density features of the fused data as input to measure the consistency of spatial gradient changes in the fused point cloud. Through an optimization process that minimizes the gradient loss, a dense 3D target model is gradually fitted. This 3D model, based on node positions, local normal vectors, and target labels as basic units, forms the basic data support for subsequent perception and path planning.

[0098] It is particularly important that step S23 includes the following:

[0099] The target semantic label data is fused with the BeiDou navigation status dataset, and a revised prediction is performed to obtain the predicted updated associated data;

[0100] Based on the prediction update correlation data, covariance correction is performed to obtain the ship covariance prediction data;

[0101] The vessel covariance prediction data is subjected to the vessel's own motion drift through the Kalman filter with a prediction period of 100ms to obtain the environment perception matrix.

[0102] In this embodiment of the present invention, data fusion is performed using target semantic label data and a Beidou navigation state dataset as input sources. This fusion process not only involves spatial alignment and temporal synchronization of information but also dynamically updates data subject to noise, uncertainty, or delay through a modified prediction mechanism, thereby generating predicted and updated associated data. This dataset incorporates multi-dimensional features such as the vessel's current position, heading, and speed, as well as the categories and spatial distribution of surrounding targets. Secondly, based on the resulting predicted and updated associated data, the system performs a covariance matrix correction, re-quantifying and weighting the uncertainties of observations from different sources at the data level to obtain corrected vessel covariance prediction data. This process effectively reduces systematic and random errors in the fused information, resulting in more accurate state estimation. Subsequently, a standard prediction-update recursion is performed using the vessel covariance prediction data as input by introducing a Kalman filter with a prediction period of 100ms. In the prediction phase, the next state is extrapolated based on the vessel's kinematic model (e.g., a constant-speed or accelerated navigation model). In the update phase, the predicted value is corrected based on new observations, and the reliability of the current estimate is evaluated based on the covariance. Ultimately, the Kalman filter outputs an environmental perception matrix, which systematically records the vessel's dynamic state, including position, velocity, and heading angle, along with its corresponding error statistics. Throughout this process, data processing goes beyond simple concatenation or averaging. Instead, targeted correction prediction and covariance control, combined with Kalman filtering based on the minimum mean square error principle, ensure the environmental perception matrix is ​​both real-time and statistically optimal, providing stable and accurate fundamental data support for subsequent navigation decisions, obstacle avoidance planning, and formation coordination.

[0103] Preferably, step S3 of channel coding the environment sensing matrix and performing subcarrier allocation includes the following:

[0104] Inputting the environment perception matrix into the first convolutional encoder to generate a coding polynomial to obtain the environment perception first convolutional code;

[0105] Using an interleaver to block-interleave the first environmental-aware convolutional code to obtain environmental-aware interleaver data;

[0106] Inputting the environment-aware interleaver data into the second encoder for convolution to obtain an environment-aware parallel convolutional code;

[0107] Subcarrier allocation is performed on the environment-aware parallel convolutional code to obtain a multi-carrier modulated signal, wherein the subcarrier allocation on the environment-aware parallel convolutional code includes the following steps:

[0108] The environment-aware parallel convolutional code adopts 64QAM SNR subcarrier modulation to obtain the environment 64-bit SNR subcarrier signal data; the environment-aware parallel convolutional code adopts 16QAM SNR subcarrier modulation to obtain the environment 16-bit subcarrier signal data; the environment-aware parallel convolutional code adopts QPSK SNR subcarrier modulation to obtain the environment QPSK subcarrier signal data.

[0109] In an embodiment of the present invention, an environment-aware matrix is ​​input as input and fed into a first convolutional encoder for processing. Convolutional encoding is performed using a set generator polynomial, wherein each bit sequence generates a convolutional codeword through state transition and output rule mapping, thereby obtaining an environment-aware first-path convolutional code. Subsequently, to enhance the randomness and burst error resistance of the encoded data, an interleaver with a block size of 1024 is used to perform a block interleaving operation on the environment-aware first-path convolutional code. The bit order is rearranged by row-column transformation and position shuffling to generate environment-aware interleaver data. Thereafter, the interleaved data is input into a second convolutional encoder for further convolutional encoding. A parallel concatenated structure is used to achieve coding gain superposition, and an environment-aware parallel concatenated convolutional code is output. The parallel concatenated convolutional structure consists of two groups of convolutional encoders with the same or different constraint lengths and generator polynomials, which are connected through an external interleaver to significantly increase the free distance of the sequence. Subsequently, for the context-aware parallel convolutional code, subcarriers are allocated according to the channel subcarrier signal-to-noise ratio (SNR) classification strategy to form a multicarrier modulated signal. Specifically, high SNR subcarriers are first mapped to 64QAM (Quadrature Amplitude Modulation) constellation points to generate high-SNR subcarrier signal data, with each symbol carrying 6 bits of information. For medium SNR subcarriers, 16QAM modulation is used to generate low-SNR subcarrier signal data, with each symbol carrying 4 bits of information. For low SNR subcarriers, QPSK (Quadrature PhaseShift Keying) modulation is used to generate low-SNR subcarrier signal data, with each symbol carrying 2 bits of information. These three different modulation schemes are dynamically adapted based on subcarrier channel quality to ensure a balance between link stability and data rate. During the entire subcarrier allocation process, subcarriers are grouped according to the real-time measured SNR threshold and resource mapping is performed according to ratio and priority. The final output is a multicarrier modulated signal that conforms to the Orthogonal Frequency Division Multiplexing (OFDM) structure, laying the foundation for subsequent data packet encapsulation and wireless transmission.

[0110] Preferably, step S3 of encapsulating IP data packets according to the multi-carrier modulated signal and relaying the IP data packets by the wireless broadband ad hoc network radio station includes the following:

[0111] The multi-carrier modulation signal is divided into time slots according to 10ms / time slot, and priority queue scheduling is performed to obtain priority queue management data, where the priority queue management data includes situation data, voice data and basic ship parameter data;

[0112] Preemptive access is performed on situation data, and time slot resources are directly allocated. Binary exponential backoff is performed on voice data to allocate time slot resources. The remaining time slot resources are competed for and transmitted for basic vessel parameter data.

[0113] Encapsulate IP data packets according to the priority queue management data, and append the target node ID and hop limit fields to the packet header to obtain multi-carrier encapsulated data of the vessel;

[0114] The wireless broadband ad hoc network radio is used to forward the multi-carrier encapsulated data of the ships in multiple channels according to the working frequency band of 2.4 GHz, bandwidth of 32 MHz and transmission filter of 20W. The relay nodes are selected according to the target node ID and hop limit to obtain the shared data of the ship formation situation.

[0115] In an embodiment of the present invention, a multi-carrier modulated signal is divided into time slots with 10 milliseconds as a basic time slot unit to form a time slicing structure. Subsequently, a priority queue scheduling strategy is set based on the importance of the data type, and the data is divided into three categories: situation data, voice data, and basic ship parameter data, and priority queue management data is generated. Among them, situation data adopts a preemptive access mechanism due to its extremely high real-time requirements, that is, it preferentially occupies resources in any time slot and directly allocates available time slots to ensure the lowest delay. Voice data uses a binary exponential backoff algorithm to allocate time slot resources. After detecting a conflict or contention failure, the re-contention time slot is randomly postponed according to the principle of exponential growth of the backoff number until the resource is successfully allocated, thereby ensuring the continuity of the voice stream and conflict control. Basic ship parameter data is transmitted in a competitive manner based on the remaining time slot resources, occupying unallocated time slots according to the ordinary random access principle. After priority queue scheduling is completed, various types of data are encapsulated into IP packets based on the priority queue management data. During the encapsulation process, a target node ID field is added to the IP packet header to specify the final receiving node of the data transmission. At the same time, a hop limit field (TTL, Time To Live) is set to control the maximum number of forwarding times of the packet in the network to prevent infinite routing loops, thereby generating multi-carrier encapsulated data for ships. Subsequently, wireless broadband ad hoc network radios are used for data relay transmission. The radio operating frequency band is set to 2.4 GHz, the bandwidth is 32 MHz, and the transmit power is limited to 20 W through out-of-band transmission filtering to control interference and energy leakage. During the relay process, suitable relay nodes are selected based on the target node ID field and the hop limit field carried in the data packet. In other words, only data relay links with current hop counts within the limit and capable of advancing toward the target node are selected for forwarding, forming a multi-path forwarding strategy. Finally, the ship formation situation sharing data is aggregated and broadcast, ensuring that all ships in the formation receive consistent environmental perception and navigation situation information.

[0116] Preferably, step S4 includes the following steps:

[0117] Step S41: Generate a global initial route for the shared data of the boat formation situation according to the voyage cost coefficient of 0.7 and the safety distance cost coefficient of 0.3;

[0118] Step S42: Perform local resolution screening on the global initial route to obtain a local obstacle avoidance path, wherein the evaluation indicators for local resolution screening are collision avoidance rule weight 0.5, navigation constraint weight 0.3, environmental interference weight 0.2, speed resolution 0.1 m / s, and heading angle resolution 1°;

[0119] Step S43: Marking route nodes based on the local obstacle avoidance path and generating a formation cooperative navigation instruction set.

[0120] In an embodiment of the present invention, a global initial route is generated for the shared data of the ship formation situation based on a preset range cost coefficient of 0.7 and a safety distance cost coefficient of 0.3. This process evaluates the path of the target ship in the sea area by combining the range cost and the safety distance cost, and optimizes the route planning to minimize the total range cost required during navigation and ensure the safety distance between ships. The range cost mainly considers the range calculation between ships, while the safety distance cost ensures that the ships in the formation always maintain appropriate spatial isolation to avoid collisions. Next, the global initial route is screened for local resolution using a preset distributed obstacle avoidance comprehensive evaluation function to obtain a local obstacle avoidance path. This evaluation function combines multiple factors, among which the collision avoidance rule weight is set to 0.5, the navigation constraint weight is 0.3, and the environmental interference weight is 0.2. It comprehensively considers the navigation rules, the constraints of ship speed and heading, and the impact of the external environment on route selection. The local resolution screening criteria further define the accuracy requirements for the local path, including a velocity resolution of 0.1 m / s and a heading angle resolution of 1°. These standards ensure the accuracy and practicality of path selection, particularly for navigation safety and precision requirements in complex waters. Finally, based on the local obstacle avoidance path, important nodes along the route are marked, and a set of coordinated navigation instructions for the formation is generated. This process identifies key navigation nodes (such as turning points and obstacle avoidance points) in the path and uses these nodes as the basis for guiding the coordination of the vessel formation. It generates corresponding navigation instructions for each vessel, ensuring that the vessels follow the designated route and instructions, thereby ensuring the coordinated operation and obstacle avoidance capabilities of the entire fleet.

[0121] It is particularly important that step S43 includes:

[0122] According to the attraction coefficient of 0.8 and the repulsion coefficient of 0.5, the local obstacle avoidance path is marked with waypoints for navigation, and important navigation marking nodes are obtained;

[0123] Based on the important navigation marking nodes, the distance measurement with a spacing error of less than 5m is carried out, and the instruction set is constructed to obtain the formation collaborative navigation instruction set.

[0124] In an embodiment of the present invention, based on a local obstacle avoidance path, a vector calculation of the attractive and repulsive forces is performed on each point on the path, wherein the attractive force coefficient is set to 0.8 and the repulsive force coefficient is set to 0.5. Through this potential field model, the boat is pushed toward the target direction while effectively avoiding obstacles. Then, the waypoints in the local path are screened and optimized according to the intensity of the combined effect of the attractive and repulsive forces, marking the key nodes with guiding significance during the navigation process, i.e., the important navigation marking nodes. Secondly, after obtaining the important navigation marking nodes, the system uses the spatial Euclidean distance measurement method to detect the spacing error of adjacent nodes with each node as the center at the data level, ensuring that the spacing error between adjacent important nodes is controlled within 5 meters. If the threshold is exceeded, a new intermediate node is inserted or the original node is adjusted to ensure the continuity and traceability of the route. Next, based on the screened and corrected set of important nodes, according to the sequential relationship between nodes and spatial coordinate data, combined with preset motion parameters such as speed control and heading change rules, node-by-node instruction construction is carried out. Each instruction set includes detailed information such as the target node ID, the three-dimensional coordinates of the target position, the required arrival speed, the heading angle limit, and the warning trigger conditions, thus forming a complete formation collaborative navigation instruction set. Throughout the process, data processing emphasizes potential field and force field modeling as the basis, supplemented by high-precision spatial ranging and dynamic threshold control, to ensure the rigor of the instruction set in terms of spatial rationality, temporal accessibility, and formation consistency. This provides a clear structure and complete data decision-making basis for the subsequent real-time navigation of the boat formation, dynamic path adjustment, and obstacle avoidance strategy implementation.

[0125] As an example of the present invention, refer to Figure 2 As shown, in this example, step S5 includes:

[0126] Step S51: Overlaying and displaying nautical charts according to the formation cooperative navigation instruction set to obtain a real-time route situation map;

[0127] Step S52: Performing a yaw-collision avoidance warning based on the real-time route situation map to obtain a level 3 warning signal;

[0128] Step S53: Based on the third-level warning signal and the real-time route situation map, control instructions are issued at a baud rate of 1 Mbps and a frame interval of 10 μs to obtain a normal control instruction set; if the packet loss rate of the ad hoc network is greater than 5%, it is determined that the communication link is interrupted and a degraded control instruction set is generated;

[0129] Step S54: constructing a vessel wireless communication control report based on the normal control instruction set and the degraded control instruction set.

[0130] In an embodiment of the present invention, a real-time route situation map is generated by overlaying and displaying the nautical chart based on the formation collaborative navigation instruction set. This process uses the real-time position and heading information of the vessels to overlay and display the current navigation status of the fleet on the electronic nautical chart system, ensuring the visualization of the route and the surrounding environment so that the commanders and crew can grasp the navigation dynamics and relative position of the fleet in real time. Then, based on the real-time route situation map, dynamic monitoring is performed using the yaw-collision avoidance warning algorithm. By analyzing the relative position and heading changes between the vessels and surrounding targets, the yaw and collision risks are promptly detected, and a three-level warning signal is generated. These warning signals are based on a collision prediction model, and are comprehensively calculated according to the relative speed, heading, and distance between the vessels and surrounding targets, and a corresponding risk level is generated so that timely action can be taken to avoid collisions. Subsequently, the three-level warning signals and the real-time route situation map are combined for data transmission via a wireless ad hoc network. Specifically, under the conditions of a set baud rate of 1Mbps and a frame interval of 10μs, control instructions are sent to the vessel to generate a normal control instruction set. These instruction sets include heading adjustment, speed control, and collision avoidance actions, etc., to ensure that the vessel performs safe operations according to the instructions. However, if the communication link packet loss rate exceeds 5% during data transmission, the system will detect that the link is unstable or interrupted, and trigger the generation of a degraded control instruction set. The degraded control instruction set usually includes simpler and more conservative instructions to ensure that the vessel can maintain basic navigation safety even when communication conditions are poor. Finally, the normal control instruction set and the degraded control instruction set are combined to generate a vessel wireless communication control report. This report contains the current control commands and adjustment instructions for all vessels, which are used to guide crew members and automation systems to implement appropriate navigation and obstacle avoidance strategies.

[0131] Preferably, step S52 includes the following:

[0132] Based on the real-time route situation map, the system makes a deviation judgment and performs CPA calculation to obtain the vessel's CPA data. Based on the real-time route situation map, the system makes a collision avoidance prediction and performs TCPA calculation to obtain the vessel's TCPA data. When the CPA data is greater than 0.5 nautical miles or the TCPA data is greater than 300s, it is judged as a green normal signal.

[0133] When the CPA data is between 0.3 nautical miles and 0.5 nautical miles and the TCPA data is between 100s and 300s, it is judged as a yellow caution signal;

[0134] When the CPA data is less than 0.3 nautical miles or the TCPA data is less than 300s, it is judged as an emergency red signal.

[0135] In this embodiment of the present invention, the calculation of the yaw-collision avoidance warning based on a real-time route situation map primarily relies on two key data points: CPA (Closest Point of Contact) and TCPA (Closest Point of Contact Time). CPA indicates the minimum contact distance between the vessel and surrounding objects, while TCPA indicates the shortest possible time to collision. These data can be used to assess the potential collision risk between the vessel and other objects. The calculation process first calculates CPA and TCPA data using real-time vessel position, speed, heading, and other dynamic information about other objects, combined with a ship dynamics model. Next, the CPA and TCPA data are classified and judged based on set thresholds to determine the vessel's current navigation safety status. When the CPA value is greater than 0.5 nautical miles and the TCPA value is greater than 300 seconds, the risk of contact with surrounding objects is very low. Therefore, a green signal is assigned, indicating safe navigation and no collision avoidance measures are required. If the CPA data is between 0.3 nautical miles and 0.5 nautical miles, and the TCPA data is between 100 seconds and 300 seconds, it indicates that there is a certain risk of approaching, and you should pay attention and take appropriate preventive measures at this time, so it is judged as a yellow caution signal. When the CPA data is less than 0.3 nautical miles, or the TCPA data is less than 300 seconds, it means that the risk of the boat approaching the target is extremely high, and emergency collision avoidance measures must be taken immediately. At this time, it is judged as a red emergency signal. Through these early warning signals, the safety situation of the boat can be monitored in real time, and data support and decision-making basis can be provided for subsequent navigation adjustments or emergency collision avoidance.

[0136] In this specification, an intelligent wireless communication control system for a boat is provided, which is used to execute the above-mentioned intelligent wireless communication control method for a boat. The intelligent wireless communication control system for a boat includes:

[0137] The data standardization module is used to collect radar raw point cloud data, AIS ship attribute data set, Beidou navigation status data set and night vision target feature data through multi-dimensional sensors to construct multi-source standardized pre-processed data of ships;

[0138] The 3D modeling and environmental perception module is used to perform 3D spatial mapping on the multi-source standardized pre-processed data of the vessel and build a target 3D model; and to construct an environmental perception matrix based on the target 3D model.

[0139] The communication coding and relay transmission module is used to perform channel coding on the environment perception matrix and perform subcarrier allocation to obtain a multi-carrier modulated signal; IP data packets are encapsulated based on the multi-carrier modulated signal and relayed by a wireless broadband ad hoc network radio to obtain shared data on the situation of the ship formation;

[0140] The route generation and collaborative control module is used to generate the global initial route based on the shared data of the boat formation situation and construct the local obstacle avoidance path; mark the important nodes of the route according to the local obstacle avoidance path and generate the formation collaborative navigation instruction set;

[0141] The situation display and warning control module is used to overlay and display nautical charts according to the formation collaborative navigation instruction set to obtain a real-time route situation map; perform deviation-collision avoidance warning on the real-time route situation map to obtain a three-level warning signal; and issue control instructions based on the three-level warning signal and the real-time route situation map to obtain a wireless communication control report for the vessel.

[0142] The beneficial effect of the present invention is that it effectively realizes the intelligent wireless communication and navigation control of the ship formation in a complex environment. Specifically, a multi-dimensional sensor is used to synchronously collect radar raw point cloud data, AIS ship attribute data, Beidou navigation status data and night vision target feature data. Through standardized preprocessing, a consistent expression framework for multi-source heterogeneous data is established to provide a unified data basis for subsequent perception and decision-making; three-dimensional space mapping is performed based on multi-source standardized data, a three-dimensional model of the target is constructed, and an environmental perception matrix is ​​generated, which improves the spatial resolution and dynamic response capability of environmental modeling, and can accurately depict the three-dimensional situation of targets and obstacles around the ship; channel coding is performed on the environmental perception matrix, and combined with a multi-carrier modulation mechanism, the subcarrier modulation method is adaptively selected according to the channel characteristics to generate a multi-carrier modulation signal, and then the IP data packet encapsulation is completed, and the wireless broadband self-organizing network radio is used in the 2.4GHz frequency band and 32MHz bandwidth conditions. Relay transmission is carried out to achieve low-latency and high-reliability situation data sharing between ships; based on the environmental perception data shared by the formation, the global initial route is generated by jointly optimizing the range cost and the safety distance cost, and the distributed obstacle avoidance comprehensive evaluation mechanism is used to perform local path correction. At the same time, important nodes of the route are marked, and a formation collaborative navigation instruction set is generated, which effectively improves the dynamic adaptability and collaborative consistency of the formation navigation; in step S5, deviation and collision avoidance warnings are carried out based on the real-time route situation map, and three-level warning classification is achieved through the CPA (closest point distance) and TCPA (closest point arrival time) indicators. Normal or degraded control instructions are dynamically issued according to the warning level and the communication link status, and finally a ship wireless communication control report is generated to ensure that the system can still maintain minimum safety control under conditions of abnormal communication link or environmental interference. Through the above-mentioned technical path, the present invention realizes the functions of multi-source heterogeneous data fusion, dynamic channel adaptive modulation, distributed relay communication guarantee, real-time route dynamic planning and intelligent command generation, etc., and solves the problems existing in the existing ship communication control technology, such as high data isolation, large transmission delay, navigation path rigidity, slow obstacle avoidance response and insufficient formation coordination, and significantly improves the communication stability, navigation safety and overall combat effectiveness of the ship formation in a complex dynamic environment.

[0143] The present invention is therefore intended to be illustrative and non-restrictive in all respects, with the scope of the invention being defined by the appended claims rather than the foregoing description, and all changes that come within the meaning and range of equivalents of the application documents are intended to be embraced therein.

[0144] The foregoing description is intended only to provide specific embodiments of the present invention, which will enable those skilled in the art to understand and implement the present invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention is not intended to be limited to the embodiments shown herein, but is to be construed in the widest possible manner consistent with the principles and novel features disclosed herein.

Claims

1. A method for controlling intelligent wireless communication of a vessel, characterized in that: The following steps are involved: Step S1: Collect radar raw point cloud data, AIS ship attribute data set, Beidou navigation status data set and night vision target feature data through multi-dimensional sensors to construct multi-source standardized pre-processed data of ships; Step S2: mapping the multi-source standardized pre-processed data of the vessel into three-dimensional space and constructing a target three-dimensional model; Construct an environmental perception matrix based on the target 3D model; Step S3: Channel coding is performed on the environment perception matrix, and subcarrier allocation is performed to obtain a multi-carrier modulated signal; IP data packets are encapsulated according to the multi-carrier modulated signal, and relayed and transmitted by a wireless broadband ad hoc network radio station to obtain shared data on the situation of the ship formation; The step of performing channel coding on the environment sensing matrix and performing subcarrier allocation to obtain a multicarrier modulated signal includes: Inputting the environment perception matrix into the first convolutional encoder to generate a coding polynomial to obtain the environment perception first convolutional code; Using an interleaver to block-interleave the first environmental-aware convolutional code to obtain environmental-aware interleaver data; Inputting the environment-aware interleaver data into the second encoder for convolution to obtain an environment-aware parallel convolutional code; The environment-aware parallel convolutional code includes a high SNR subcarrier, a medium SNR subcarrier, and a low SNR subcarrier; For the environment-aware parallel convolutional code, subcarriers are allocated according to the channel subcarrier signal-to-noise ratio classification strategy to form a multi-carrier modulation signal, including: using 64QAM SNR subcarrier modulation for high signal-to-noise ratio subcarriers to obtain 64-bit environmental SNR subcarrier signal data; using 16QAM SNR subcarrier modulation for medium signal-to-noise ratio subcarriers to obtain 16-bit environmental subcarrier signal data; using QPSK SNR subcarrier modulation for low signal-to-noise ratio subcarriers to obtain environmental QPSK subcarrier signal data; The encapsulation of IP data packets according to the multi-carrier modulation signal and the relay transmission by the wireless broadband ad hoc network radio station include the following: The multi-carrier modulation signal is divided into time slots according to 10ms / time slot, and priority queue scheduling is performed to obtain priority queue management data, where the priority queue management data includes situation data, voice data and basic ship parameter data; Preemptive access is performed on situation data, and time slot resources are directly allocated. Binary exponential backoff is performed on voice data to allocate time slot resources. The remaining time slot resources are competed for and transmitted for basic vessel parameter data. Encapsulate IP data packets according to the priority queue management data, and append the target node ID and hop limit fields to the packet header to obtain multi-carrier encapsulated data of the vessel; The wireless broadband ad hoc network radio is used to forward the multi-carrier encapsulated data of the ships in a multi-hop manner according to the working frequency band of 2.4GHz, bandwidth of 32MHz, and transmission power of 20W. The relay nodes are selected according to the target node ID and hop limit to obtain the shared data of the ship formation situation. Step S4: Generate a global initial route based on the shared data of the boat formation situation and construct a local obstacle avoidance path; mark the route nodes according to the local obstacle avoidance path and generate a formation collaborative navigation instruction set; Step S5: The nautical chart is superimposed and displayed according to the formation collaborative navigation instruction set to obtain a real-time route situation map; a deviation-collision avoidance warning is performed on the real-time route situation map to obtain a three-level warning signal; a control instruction is issued to the central control system based on the three-level warning signal and the real-time route situation map to obtain a vessel wireless communication control report.

2. The method for controlling intelligent wireless communication for ships and boats according to claim 1, characterized in that: Step S1 includes the following steps: Step S11: deploying navigation radars on the deck and mast of the vessel, operating in the X-band frequency band, with a detection range greater than 20 nautical miles and a resolution less than 30 meters, to scan the surrounding sea area, collecting target azimuth, distance, speed, and projected area data to obtain radar raw point cloud data; Step S12: Using a one-way AIS receiver with a frequency of 161.975 MHz and a decoding rate of 1 second, the MMSI number, speed, heading, and static information of the ship are collected to generate an AIS ship attribute data set. Step S13: Installing a visible light + infrared dual-band night vision device on the front of the boat to collect optical images in night or low-light environments, and extracting target outline, type and relative distance data to obtain night vision target feature data; Step S14: Using the BeiDou-3 navigation terminal, the vessel's latitude and longitude, speed, and attitude angle data are collected in real time to generate a BeiDou navigation status data set; Step S15: Check the radar point cloud data for a false alarm rate less than or equal to The sea clutter is suppressed, the timeout and unupdated signals of the AIS ship attribute data set are removed, and the night vision target feature data with a PSNR greater than 35dB are defogged and motion blur compensated to obtain the multi-source standardized preprocessing data of ships.

3. The method for controlling intelligent wireless communication for ships and boats according to claim 1, wherein: Step S2 includes the following steps: Step S21: mapping the multi-source standardized pre-processed data of the vessel into a three-dimensional space and constructing a target three-dimensional model; Step S22: Outputting target semantic label data using the target three-dimensional model; Step S23: The target semantic label data is fused with the Beidou navigation status data set, and the vessel's own motion drift is performed through a Kalman filter with a prediction period of 100ms to obtain an environmental perception matrix.

4. The method for controlling intelligent wireless communication for ships and boats according to claim 3, characterized in that: Step S21 includes the following steps: Step S211: Convert the polar coordinates of the radar point cloud data into the local radar ENU coordinate system, and obtain WGS-84 radar geographic coordinates through coordinate rotation and translation; perform WGS-84 coordinate system conversion based on the AIS ship attribute dataset to obtain WGS-84AIS coordinates; extract camera calibration parameters based on the night vision target feature data, and back-project the pixel coordinates into the local 3D coordinate system for local WGS-84 conversion to obtain WGS-84 night vision device coordinates; Step S212: Distance weights are assigned according to the weights of WGS-84 radar geographic coordinates being 0.6, WGS-84 AIS coordinates being 0.3, and WGS-84 night vision device coordinates being 0.1, and the data are mapped to three-dimensional space to obtain vessel three-dimensional space coordinate fusion data; Step S213: Screen the fused data based on the three-dimensional spatial coordinate fusion data of the vessel with a position error less than 2m, and construct the model using a preset gradient loss function to obtain a target three-dimensional model.

5. The method for controlling intelligent wireless communication for ships and boats according to claim 1, characterized in that: Step S4 includes the following steps: Step S41: Generate a global initial route for the shared data of the boat formation situation according to the voyage cost coefficient of 0.7 and the safety distance cost coefficient of 0.3; Step S42: Perform local resolution screening on the global initial route to obtain a local obstacle avoidance path, wherein the evaluation indicators for local resolution screening are collision avoidance rule weight 0.5, navigation constraint weight 0.3, environmental interference weight 0.2, speed resolution 0.1 m / s, and heading angle resolution 1°; Step S43: Marking route nodes based on the local obstacle avoidance path and generating a formation cooperative navigation instruction set.

6. The method for controlling intelligent wireless communication for ships and boats according to claim 1, characterized in that: Step S5 includes the following steps: Step S51: Overlaying and displaying nautical charts according to the formation cooperative navigation instruction set to obtain a real-time route situation map; Step S52: Performing a yaw-collision avoidance warning based on the real-time route situation map to obtain a level 3 warning signal; Step S53: Based on the third-level warning signal and the real-time route situation map, control instructions are issued at a baud rate of 1 Mbps and a frame interval of 10 μs to obtain a normal control instruction set; if the packet loss rate of the ad hoc network is greater than 5%, it is determined that the communication link is interrupted and a degraded control instruction set is generated; Step S54: constructing a vessel wireless communication control report based on the normal control instruction set and the degraded control instruction set.

7. The method for controlling intelligent wireless communication for ships and boats according to claim 6, characterized in that: Step S52 includes the following: Based on the real-time route situation map, the ship can make a yaw judgment and perform CPA calculation to obtain the ship's CPA data; based on the real-time route situation map, the ship can make a collision avoidance prediction and perform TCPA calculation to obtain the ship's TCPA data; When the CPA data is greater than 0.5 nautical miles or the TCPA data is greater than 300s, it is judged as a green normal signal; When the CPA data is between 0.3 nautical miles and 0.5 nautical miles and the TCPA data is between 100s and 300s, it is judged as a yellow caution signal; When the CPA data is less than 0.3 nautical miles or the TCPA data is less than 300s, it is judged as an emergency red signal.

8. An intelligent wireless communication control system for boats, characterized in that: The method for controlling the intelligent wireless communication of a ship according to claim 1 is used to execute the method, which comprises: The data standardization module is used to collect radar raw point cloud data, AIS ship attribute data set, Beidou navigation status data set and night vision target feature data through multi-dimensional sensors to construct multi-source standardized pre-processed data of ships; The 3D modeling and environmental perception module is used to perform 3D spatial mapping on the multi-source standardized pre-processed data of the vessel and build a target 3D model; and to construct an environmental perception matrix based on the target 3D model. The communication coding and relay transmission module is used to perform channel coding on the environment perception matrix and perform subcarrier allocation to obtain a multi-carrier modulated signal; IP data packets are encapsulated based on the multi-carrier modulated signal and relayed by a wireless broadband ad hoc network radio to obtain shared data on the situation of the ship formation; The route generation and collaborative control module is used to generate the global initial route based on the shared data of the boat formation situation and construct the local obstacle avoidance path; mark the important nodes of the route according to the local obstacle avoidance path and generate the formation collaborative navigation instruction set; The situation display and warning control module is used to overlay and display nautical charts according to the formation collaborative navigation instruction set to obtain a real-time route situation map; perform deviation-collision avoidance warning on the real-time route situation map to obtain a three-level warning signal; and issue control instructions based on the three-level warning signal and the real-time route situation map to obtain a wireless communication control report for the vessel.

Images