Collaborative navigation data management method between marine platform and unmanned boat

Abstract

The present invention discloses a collaborative navigation data management method between an ocean platform and an unmanned boat, and specifically relates to the technical field of collaborative navigation data management, including the following steps: by real-time collection of ocean environment data, combined with ocean platform, unmanned boat operation data and collaborative navigation task data, various data sets are generated and collaborative analysis is performed to obtain a computing performance index and a task complexity index. Fuzzy reasoning is used to dynamically determine the data processing location, i.e., an ocean platform, an unmanned boat or the cloud, and the ocean environment data is processed at a selected location to generate an optimized data set for navigation path planning. At the same time, the communication frequency of the communication path is dynamically adjusted based on the computing performance index and the task complexity index to ensure efficient resource allocation and real-time execution of tasks, thereby improving the intelligence level of collaborative navigation; the present invention can not only improve the efficiency and reliability of the overall collaboration between the ocean platform and the unmanned boat, but also adjust the communication frequency to ensure task requirements while saving communication resources.

Description

Technical Field

[0001] The present invention relates to the technical field of collaborative navigation data management, and more specifically, to a collaborative navigation data management method between an ocean platform and an unmanned boat. Background Art

[0002] As important components of ocean exploration and monitoring, marine platforms and unmanned boats each have unique functions and advantages. In actual missions, the efficient collaboration between the two can significantly improve the efficiency and effectiveness of mission execution. Marine platforms are usually used as fixed or semi-fixed offshore operation stations, usually equipped with powerful computing resources, suitable for processing complex data calculations and mission planning, and can simultaneously collect multiple types of marine environmental data (such as ocean currents, temperature, weather, etc.), providing rich data support for missions.

[0003] An unmanned boat is a small, autonomous or remotely controlled marine device that can move flexibly in the marine environment and adapt to dynamic mission requirements. Its proximity to the marine environment enables it to quickly acquire and process local data, but due to its small size, relatively limited computing power and energy reserves, it is difficult to cope with complex or long-term missions.

[0004] The collaborative relationship between the ocean platform and the unmanned boat is the key to achieving efficient navigation and data management. The ocean platform provides powerful computing power and energy support, and the unmanned boat provides flexible mobility and real-time perception capabilities. The two can make up for their respective shortcomings through collaboration, but the computing task allocation of the ocean platform and the unmanned boat fails to fully utilize their respective advantages, and the data processing and path planning are not dynamic enough, making it difficult to adapt to the changing ocean environment, affecting the collaborative efficiency. Summary of the invention

[0005] To achieve the above object, the present invention provides the following technical solutions:

[0006] The collaborative navigation data management method between the marine platform and the unmanned boat includes the following steps:

[0007] The marine environment data is collected in real time to obtain a marine environment data set, and the marine platform operation data, the unmanned boat operation data and the collaborative navigation task data are obtained to obtain a marine platform operation data set, an unmanned boat operation data set and a collaborative navigation task data set;

[0008] Based on the marine platform operation data set, the unmanned boat operation data set and the collaborative navigation task data set, a collaborative analysis operation is performed to generate a computing performance index and a task complexity index, and then the computing performance index and the task complexity index are fuzzy reasoned together, and the location of the marine environment data set for data processing is determined according to the result of the fuzzy reasoning;

[0009] At the position corresponding to the result of fuzzy reasoning, data processing is performed on the marine environment data set to obtain a processed data set, and navigation path planning is performed according to the processed data set;

[0010] Based on the computing performance index and the task complexity index, the communication frequency of the communication path is optimized and adjusted.

[0011] In a preferred embodiment, processing the marine environment data set refers to intelligently classifying and fusing various marine environment data in the marine environment data set according to a preset processing strategy.

[0012] In a preferred embodiment, when performing navigation path planning, a graph-based path planning algorithm, an optimization-based path planning algorithm, and a deep learning-based path planning algorithm are selected, and one of the planning algorithms is used.

[0013] In a preferred embodiment, the acquisition logic of the calculation performance index is:

[0014] Obtain the computing performance of the ocean platform and mark it as Pplatform, the computing performance of the unmanned boat and mark it as Pdrone, the computing load occupancy rate of the ocean platform and mark it as Lplatform, the computing load occupancy rate of the unmanned boat and mark it as Ldrone, and substitute them into the following calculation formula:

[0015] ; represents the computing performance index, represents the dynamic load adjustment factor, represents the resource balancing factor;

[0016] The resource balancing factor is calculated as follows: ;

[0017] Dynamic load adjustment factor Obtained through the following calculation:

[0018] ; A constant to prevent the denominator from being zero.

[0019] In a preferred embodiment, the logic for obtaining the task complexity index is:

[0020] The entire navigation area is divided into multiple sub-areas, and the historical data and real-time data of the ocean current speed in each sub-area are obtained. For each sub-area, the absolute value of the difference between the current ocean current speed and the ocean current speed value in the previous time window is calculated first, and then the absolute value is divided by the ocean current speed value in the previous time window to obtain the change amplitude of the ocean current speed. Then, the change amplitude values ​​of all sub-areas are weighted averaged to obtain the environmental complexity coefficient of the entire navigation area. When calculating the weighted average, the weight coefficient of each sub-area is defined as the reciprocal of the square of the Euclidean distance between the coordinates of the center point of the sub-area and the current coordinates of the unmanned boat;

[0021] Divide the number of obstacles in the entire navigation area by the entire navigation area to obtain the obstacle complexity coefficient;

[0022] Get the straight-line distance between the current coordinates of the unmanned boat and the destination coordinates and mark it as LL, the number of waypoints that need to be passed in the entire navigation area and mark it as TJ, the preset path planning tolerance update delay time value and mark it as YC, and then calculate:

[0023] ; Represents the task tolerance coefficient, w1 and w2 are both preset proportional coefficients;

[0024] Finally, the task tolerance coefficient , obstacle complexity coefficient, and environment complexity coefficient are substituted into the pre-trained neural network model, and the output result of the neural network model is the task complexity index.

[0025] In a preferred embodiment, the logic of fuzzy reasoning is:

[0026] The computing performance index and the task complexity index are taken as input variables, and the location for data processing is taken as the output variable. The input variables are fuzzified and the values ​​of the input variables are converted into fuzzy sets. The output variables are fuzzified and the output variables are converted into fuzzy sets. Fuzzy rules are formulated to describe the degree of data processing adaptation under different data type combinations. The fuzzified input variables are inferred through fuzzy rules to obtain the location for data processing of the marine environment dataset. The location for data processing of the marine environment dataset is one of the marine platform, unmanned boat, and cloud.

[0027] In a preferred embodiment, optimizing and adjusting the communication frequency of the communication path based on the computing performance index and the task complexity index refers to:

[0028] If the location for processing the marine environment dataset is an offshore platform or unmanned boat, make the following adjustments:

[0029] ; The task complexity index is is the basic communication frequency, is the preset adjustment coefficient 1, Indicates the communication frequency of the communication link between the ocean platform and the unmanned boat after adjustment;

[0030] If the location for processing the marine environment dataset is the cloud, make the following adjustments:

[0031] ; is the preset task quality standard value, is the preset adjustment factor 2, It indicates the communication link between the marine platform and the cloud, and the communication link between the unmanned boat and the cloud after adjustment.

[0032] Technical effects and advantages of the present invention:

[0033] The present invention realizes the dynamic allocation of data processing tasks by comprehensively analyzing the computing performance and task complexity index of the marine platform and the unmanned boat, giving full play to the computing power of the marine platform and the real-time advantages of the unmanned boat, avoiding resource waste, and improving the efficiency and reliability of overall coordination. The location of data processing (marine platform, unmanned boat or cloud) is dynamically determined by the fuzzy reasoning method, and the data processing strategy can be adjusted according to the real-time task requirements and equipment status, so as to adapt to the complex and changeable marine environment and meet the needs of high real-time tasks. The present invention proposes a dynamic adjustment mechanism of communication frequency based on the task complexity index and the computing performance index. According to the complexity of the task and the computing power of the equipment, the mechanism intelligently adjusts the communication frequency between the marine platform and the unmanned boat, as well as the communication frequency with the cloud, which not only guarantees the task requirements but also saves communication resources.

[0034] By achieving efficient collaboration between the marine platform and the unmanned boat, the present invention solves the shortcomings of the traditional single-device processing mode. The task division among multiple devices is clear, resources are complementary, and collaborative work significantly improves the efficiency and flexibility of task execution. After the data processing is completed, the present invention can select the optimal path planning algorithm (graph-based algorithm, optimization algorithm or deep learning algorithm) according to the fused data set to generate an efficient and reliable navigation path. The dynamic adjustment capability of path planning enhances the adaptability and navigation accuracy of the unmanned boat in a complex marine environment.

[0035] The present invention constructs a quantitative model of multi-dimensional task difficulty through a comprehensive evaluation of the task complexity index, combined with environmental complexity, obstacle complexity and task tolerance, and introduces a neural network model for optimization, which significantly improves the decision-making ability in complex task scenarios. Dynamic optimization of task allocation and communication frequency not only reduces total energy consumption, but also reduces the waste of communication resources, which meets the requirements of modern ocean monitoring for low energy consumption and high efficiency. By dynamically adjusting and optimizing data processing and communication links, the present invention can effectively reduce the risk of mission failure caused by uneven load or communication delay in the system, and provide higher reliability and stability for the navigation mission of unmanned boats. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] In order to facilitate understanding by those skilled in the art, the present invention is further described below in conjunction with the accompanying drawings;

[0037] Figure 1 It is a schematic diagram of the collaborative navigation data management method between the ocean platform and the unmanned boat in the present invention. DETAILED DESCRIPTION

[0038] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only 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 ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0039] Reference Figure 1 The following embodiments are obtained:

[0040] Embodiment 1: A collaborative navigation data management method between an ocean platform and an unmanned boat comprises the following steps:

[0041] Collect marine environment data in real time to obtain marine environment data sets. At the same time, obtain marine platform operation data, unmanned boat operation data and collaborative navigation task data to obtain marine platform operation data sets, unmanned boat operation data sets and collaborative navigation task data sets. Real-time collection of marine environment data is the basis of collaborative navigation, ensuring that real-time change information such as ocean currents, weather, and temperature can be obtained to provide support for navigation decisions. Collect the operating status of marine platforms and unmanned boats (such as computing power, load rate, etc.) separately to provide a basis for subsequent judgment of data processing locations and task scheduling. Collaborative navigation task data (such as navigation path planning goals and task real-time requirements) are used to comprehensively analyze task requirements to ensure the pertinence and accuracy of system decisions. Different types of data are classified and stored to form multiple independent data sets for subsequent analysis and decision-making operations.

[0042] Based on the marine platform operation data set, unmanned boat operation data set and collaborative navigation task data set, collaborative analysis operations are performed to generate computing performance index and task complexity index, and then fuzzy reasoning is performed on the computing performance index and task complexity index together. According to the results of fuzzy reasoning, the location of data processing in the marine environment data set is determined; by comprehensively analyzing the operation data (computing performance, load rate, etc.) and task requirements (task complexity, real-time requirements, etc.), computing performance index and task complexity index are generated to quantify the matching degree between equipment capabilities and task difficulty, ensure that data analysis is scientific and systematic, and provide a reliable basis for subsequent decision-making. Fuzzy reasoning infers the optimal location for data processing (marine platform, unmanned boat or cloud) based on computing performance index and task complexity index. The adaptability in different scenarios is described by fuzzy rules, and changes in task requirements and operating environment are flexibly responded to. The purpose of judging the data processing location is to optimize resource utilization (such as computing performance and communication frequency), while meeting the real-time and complexity requirements of tasks, and improving the overall efficiency of collaborative navigation.

[0043] At the location corresponding to the result of fuzzy reasoning, the marine environment data set is processed to obtain a processed data set, and navigation path planning is performed based on the processed data set; based on the result of fuzzy reasoning, the marine environment data is processed at the most suitable location (marine platform, unmanned boat or cloud) to ensure processing efficiency and real-time performance. Through the preset intelligent classification and fusion strategy, the key information in the marine environment data is extracted and sorted to generate an optimized data set for navigation path planning. Using the processed data set, a path planning algorithm (graph-based, optimization or deep learning algorithm) suitable for the task requirements is selected to generate the optimal navigation path for the unmanned boat. The planning algorithm designs the route according to the task objectives (such as the shortest path, obstacle avoidance or energy optimization) to ensure the safety of the unmanned boat in the dynamic marine environment and the efficiency of task completion.

[0044] It should be noted that the location of data processing (on the ocean platform, unmanned boat or in the cloud) is determined after fuzzy reasoning. Before the data processing location is determined, the data has not been classified and fused, so it is not possible to directly use the unprocessed data to complete complex calculations, such as partition analysis or obstacle statistics.

[0045] In order to avoid logical conflicts, the division of responsibilities before and after data processing is clarified here:

[0046] Before data processing (before finding the processing location): In the fuzzy reasoning stage, the data that needs to be relied on should be basic, coarse-grained statistical information, rather than the results of complex processing. It can be adjusted to the following logic: Use the original statistical information of the marine environment data, such as: the maximum, minimum, and standard deviation of the ocean current speed. The preliminary detection number of obstacles and the rough boundary range of the navigation area. Based on these simple statistical information, in conjunction with the operating data sets and task data sets of the marine platform and unmanned boat, calculate the input required for fuzzy reasoning (such as the calculation performance index and the task complexity index). The fuzzy reasoning stage can only use simple statistical information, such as the original ocean current speed, the preliminary detection number of obstacles, etc., which do not rely on intelligent classification and fusion.

[0047] After data processing (after determining the processing location): After the processing location is determined, data processing is performed, including: Intelligent classification: Detailed division of marine environmental data by category (such as ocean current data, obstacle data). Intelligent fusion: Integrate multi-source data according to task requirements to generate optimized data sets, such as: detailed navigation area partitioning; accurate obstacle distribution statistics; local trend analysis of ocean current changes. Based on the optimized data set, use a path planning algorithm (graph algorithm, optimization algorithm or deep learning algorithm) to generate the optimal path.

[0048] Based on the computing performance index and the task complexity index, the communication frequency of the communication path is optimized and adjusted. The communication frequency determines the efficiency of data interaction between the marine platform, the unmanned boat and the cloud. In high-real-time tasks, too low a frequency will lead to decision delays, while too high a frequency may cause bandwidth waste or link congestion. Adjusting the communication frequency not only improves resource utilization efficiency, but also enhances the adaptability of the present invention to dynamic environments, ensuring the smooth completion of navigation tasks.

[0049] Data processing of the marine environment data set refers to intelligently classifying and integrating the various marine environment data in the marine environment data set according to the preset processing strategy. The following is a specific example to illustrate how to use the preset processing strategy to complete these operations:

[0050] Scenario assumption: The task is to generate an optimized navigation path for the unmanned boat based on real-time collected ocean environment data (such as ocean current speed, temperature, obstacle data, etc.).

[0051] Intelligent data classification: Purpose of classification: Organize different types of marine environment data into categories to facilitate subsequent integration and analysis. Example data: Ocean current data: including speed, direction, and change trend. Temperature data: including water surface temperature and vertical temperature gradient. Obstacle data: including obstacle location, size, and moving speed.

[0052] Classification strategy: Rule-based classification: Use the predefined structure of the data (such as field name, data format) for direct classification. For example, data containing "velocity" in the field is classified as ocean current data, and data starting with "temperature" is classified as temperature data.

[0053] Classification based on machine learning: Use classification algorithms (such as decision trees and support vector machines) to classify unstructured data (such as sensor logs). Example: Use target detection algorithms (such as YOLO or FasterR-CNN) to identify obstacles in image data collected by a camera and extract features such as the location and size of the obstacles.

[0054] Intelligent data fusion: Purpose of fusion: Integrate multi-source data (such as data collected by different sensors at different times) to generate a more comprehensive and consistent data set. Example operation: Ocean current data fusion: Perform weighted averaging of ocean current velocity data collected by multiple sensors in the same time period to remove noise. Use the Kalman filter algorithm to predict the current missing data points based on historical ocean current velocity data.

[0055] Temperature data fusion: The temperature data collected by sensors at different depths are interpolated to generate vertical temperature profiles. Time series-based fusion techniques (such as the ARIMA model) are used to process the temporal variation of temperature data.

[0056] Obstacle data fusion: The obstacle detection data from the unmanned boat radar and camera are fused to eliminate redundant information through data association (such as similarity in position and time). Example: If the radar detects an obstacle at a certain location, and the camera finds the same object at the same location, the size information of the two can be weighted averaged to obtain a more accurate obstacle size.

[0057] Big data processing tools for classification and fusion technology tools: Apache Spark: can perform distributed classification and fusion operations on massive amounts of real-time collected data. TensorFlow / PyTorch: used for the implementation of deep learning algorithms, such as image classification and target detection. Algorithm examples: DBSCAN (density clustering): spatially clusters obstacle data and marks obstacle groups in high-density areas. Kalman filter: used for time series prediction and data smoothing of ocean current velocity data. Bayesian fusion: used for probabilistic fusion of multi-sensor data to generate comprehensive information on obstacle locations and attributes.

[0058] Comprehensive example: Complete classification and fusion process:

[0059] Classification: Get ocean current speed and direction data from sensor A, labeled "ocean current data". Get temperature data from sensor B, labeled "temperature data". Get lidar point cloud data of obstacles from sensor C, labeled "obstacle data".

[0060] Fusion: Use Kalman filtering to smooth the ocean current velocity data from sensor A and obtain the ocean current trend in each time window. Use the Bayesian fusion method to jointly estimate the position and attributes of obstacle data from the lidar and camera. Use the time series interpolation method for multi-sensor temperature data to generate a temperature profile.

[0061] Output: The integrated dataset includes precise ocean current trends, vertical temperature profiles, and the exact location and size of obstacles, which will be directly used for navigation path planning.

[0062] When performing navigation path planning, select one of the following planning algorithms: a graph-based path planning algorithm, an optimization-based path planning algorithm, or a deep learning-based path planning algorithm.

[0063] Graph-based path planning algorithms such as the A* algorithm are heuristic search algorithms that use the cost function f(n)=g(n)+h(n) for path planning, where: g(n) is the actual cost from the starting point to node n. h(n) is the estimated cost from node n to the target point (heuristic function). It is suitable for planning the shortest path when the ocean map is known, and can efficiently search for paths with lower costs. Applicable scenarios: known sea areas, detailed terrain or obstacle information, the goal is to find the shortest path and a certain degree of real-time performance is required.

[0064] Optimization-based path planning algorithms such as genetic algorithms (GA) imitate the evolutionary process of natural selection and find the optimal path through population initialization, crossover, mutation and selection. It can optimize paths in complex and nonlinear environments and can add multi-objective optimization (such as path length, energy consumption and safety). Applicable scenarios: complex sea areas, need to comprehensively optimize multiple goals (such as obstacle avoidance, energy efficiency, etc.), and offline optimization paths when time permits.

[0065] Path planning algorithms based on deep learning, such as deep reinforcement learning algorithms, combine deep learning and reinforcement learning. The model learns the optimal navigation strategy through environmental interaction. It is suitable for continuous decision-making in dynamic environments, can handle high-dimensional state spaces, and adapt to complex marine environments. Common methods: DDPG (Deep Deterministic Policy Gradient): Applicable to continuous action spaces. PPO (Proximal Policy Optimization): Applicable to mission scenarios with high dynamic changes. Applicable scenarios: The marine environment is complex and the dynamic changes are obvious (such as real-time changes in tides and ocean currents). Navigation strategies need to be trained based on a combination of historical data and real-time data.

[0066] The logic for obtaining the calculation performance index is:

[0067] Obtain the computing performance of the ocean platform and mark it as Pplatform, the computing performance of the unmanned boat and mark it as Pdrone, the computing load occupancy of the ocean platform and mark it as Lplatform, and the computing load occupancy of the unmanned boat and mark it as Ldrone. Pplatform and Pdrone represent the computing performance of the ocean platform and the unmanned boat, such as CPU performance, memory capacity and other hardware capabilities, which are used to measure the computing power of each device in processing data. The higher the computing performance, the stronger the adaptability of the device in data processing. Lplatform and Ldrone represent the current computing load occupancy of the ocean platform and the unmanned boat (ranging from 0 to 1), which are used to reflect the current workload of the device. The higher the computing load, the lower the ability of the device to process new tasks.

[0068] Substitute the following calculation formula:

[0069] ; represents the computing performance index, represents the dynamic load adjustment factor, represents the resource balancing factor;

[0070] The resource balancing factor is calculated as follows: The resource balance factor reflects the degree of balance in resource allocation between the marine platform and the unmanned boat. The greater the difference in computing performance between the platform and the unmanned boat, the higher the resource balance factor, indicating an imbalance in resource allocation. Resource balance is introduced to ensure that devices with weaker computing performance will not be overused.

[0071] The computing performance index is used to quantify the computing power compatibility of the marine platform and the unmanned boat, providing a basis for the selection of the data processing location. The computing performance index is directly used for fuzzy reasoning to determine the location of data processing (marine platform, unmanned boat or cloud). The formula comprehensively considers computing performance and load distribution through dynamic load adjustment factors and balance factors, so that tasks can be evenly distributed and single equipment can be prevented from overloading.

[0072] Dynamic load adjustment factor Obtained through the following calculation:

[0073] ; A constant set to prevent the denominator from being zero. The dynamic load adjustment factor is used to adjust the adaptability of computing performance under dynamic load changes. If the platform load is higher than the unmanned boat, the dynamic load adjustment factor increases, indicating that the platform computing performance is greatly restricted. If the unmanned boat load is higher than the platform, the dynamic load adjustment factor decreases, and the relative weight of the unmanned boat performance decreases, which can dynamically reflect the impact of load changes on the computing power of different devices.

[0074] The logic for obtaining the task complexity index is:

[0075] The entire navigation area is divided into multiple sub-areas, and the historical and real-time data of the ocean current speed in each sub-area are obtained. For each sub-area, the absolute value of the difference between the current ocean current speed and the ocean current speed value in the previous time window is calculated first, and then the absolute value is divided by the ocean current speed value in the previous time window to obtain the change amplitude of the ocean current speed. Then, the change amplitude values ​​of all sub-areas are weighted averaged to obtain the environmental complexity coefficient of the entire navigation area. When calculating the weighted average, the weight coefficient of each sub-area is defined as the reciprocal of the square of the Euclidean distance between the coordinates of the center point of the sub-area and the current coordinates of the unmanned boat; the square reciprocal of the Euclidean distance increases rapidly as the distance decreases, reflecting that the sub-area closer to the position of the unmanned boat has a greater impact on the task. Using the form of the square reciprocal, the weight decreases rapidly as the distance increases, which reasonably limits the impact of the long-distance sub-area on the overall calculation. Compared with the simple reciprocal, the square reciprocal decays faster and is more suitable for the dynamic ocean scenes faced by the navigation task.

[0076] The obstacle complexity coefficient is obtained by dividing the number of obstacles in the entire navigation area by the entire navigation area. It reflects the density of obstacles. The higher the value, the denser the obstacles in the area, and the higher the accuracy and complexity required for path planning.

[0077] Get the straight-line distance between the current coordinates of the unmanned boat and the destination coordinates and mark it as LL, the number of waypoints that need to be passed in the entire navigation area and mark it as TJ, the preset path planning tolerance update delay time value and mark it as YC, and then calculate:

[0078] ; Represents the task tolerance coefficient, which is a measure of the task's tolerance for path update delay and path planning, and reflects the task's requirements for the length, complexity, and time tolerance of the navigation path. The higher the value, the higher the task's requirements for real-time performance. w1 and w2 are both preset proportional coefficients used to balance the impact of path length and the number of waypoints.

[0079] Finally, the task tolerance coefficient , obstacle complexity coefficient, and environment complexity coefficient are substituted into the pre-trained neural network model, and the output result of the neural network model is the task complexity index. The environment complexity coefficient, obstacle complexity coefficient, and task tolerance coefficient are input into the pre-trained neural network model, and the model outputs the task complexity index. The convolutional neural network integrates the complexity of multiple dimensions into a single indicator, which is convenient for quantitative evaluation of task difficulty and provides a basis for navigation path planning and communication frequency adjustment.

[0080] The logic of fuzzy reasoning is:

[0081] The computing performance index and the task complexity index are taken as input variables, and the location for data processing is taken as the output variable. The input variables are fuzzified and the values ​​of the input variables are converted into fuzzy sets. The output variables are fuzzified and the output variables are converted into fuzzy sets. Fuzzy rules are formulated to describe the degree of data processing adaptation under different data type combinations. The fuzzified input variables are inferred through fuzzy rules to obtain the location for data processing of the marine environment dataset. The location for data processing of the marine environment dataset is one of the marine platform, unmanned boat, and cloud.

[0082] The computational performance index and the task complexity index are quantitative indicators used to describe the computing capabilities and task complexity of marine platforms and unmanned boats. These indicators may show continuous changes in actual values, so they are fuzzified into fuzzy sets. For example, the computational performance index can be divided into "high adaptability", "medium adaptability" and "low adaptability", and the task complexity index can be divided into fuzzy sets such as "simple", "medium complexity" and "complex".

[0083] Fuzzification of output variables: The location of data processing (offshore platform, unmanned boat or cloud) is also fuzzified as an output variable and processed into a fuzzy set. For example, "high adaptability of offshore platform" and "medium adaptability of unmanned boat" indicate the adaptability of data processing in different locations.

[0084] Fuzzy rules are formulated based on different combinations of computing performance index and task complexity index. For example, when the computing performance index is "high adaptability" and the task complexity index is "simple", unmanned boats are preferred for data processing. When the computing performance index is "medium adaptability" and the task complexity index is "complex", marine platforms are preferred. When the computing performance index is "low adaptability" and the task complexity index is "complex", the cloud is selected for data processing.

[0085] Fuzzy reasoning: The fuzzified input variables are reasoned through fuzzy rules to obtain the fitness of different data processing positions. The position with higher fitness is considered to be the optimal position for current data processing.

[0086] Defuzzification: Defuzzify the result of fuzzy reasoning and output a clear processing location (offshore platform, unmanned boat or cloud).

[0087] The computing power and task load of the offshore platform, unmanned boat and cloud are different. Choosing the right data processing location can avoid wasting resources. For example, when the computing power of the unmanned boat is sufficient and the task complexity is low, the processing can be completed directly on the unmanned boat without occupying the resources of the cloud or offshore platform.

[0088] The location of data processing will directly affect processing efficiency and response speed. For example, when complex tasks require high-performance computing, the cloud is preferred; and for tasks with high real-time requirements, the nearest unmanned boat can be selected for processing to reduce communication delays.

[0089] By dynamically selecting the data processing location, we can flexibly respond to environmental changes and adjustments to task complexity. For example, when the load rate of the ocean platform is high, the data processing task can be transferred to an unmanned boat or the cloud.

[0090] Advantages of offshore platforms: strong computing power and low communication latency; suitable for handling complex tasks, especially when multiple data sources need to be integrated. Applicable scenarios: When the task complexity index is high and the offshore platform load rate is low, offshore platforms are preferred for processing data.

[0091] Advantages of unmanned boats: strong real-time performance, close distance to sensors, can reduce data transmission time; suitable for handling lightweight tasks. Applicable scenarios: Unmanned boats are preferred when the task complexity index is low and the unmanned boat computing load is small.

[0092] Advantages of the cloud: Almost unlimited computing power, suitable for handling large-scale complex tasks. Disadvantages: High communication latency, dependent on stable network connection. Applicable scenarios: Choose the cloud when the task is complex and the computing power of the offshore platform and unmanned boat is insufficient.

[0093] Based on the computing performance index and the task complexity index, optimizing the communication frequency of the communication path means:

[0094] If the location for processing the marine environment dataset is an offshore platform or unmanned boat, make the following adjustments:

[0095] ; The task complexity index is is the basic communication frequency, is the preset adjustment coefficient 1, Indicates the communication frequency of the communication link between the ocean platform and the unmanned boat after adjustment;

[0096] If the location for processing the marine environment dataset is the cloud, make the following adjustments:

[0097] ; is the preset task quality standard value, is the preset adjustment factor 2, It indicates the communication link between the marine platform and the cloud, and the communication link between the unmanned boat and the cloud after adjustment.

[0098] The core purpose of adjusting the communication frequency is to dynamically adjust the efficiency of the communication link according to the task requirements (task complexity index) and the computing performance index to meet the real-time and complexity requirements of the task while avoiding resource waste. The formula implements dynamic adjustment through the following logic: the more complex the task, the higher the communication requirements, and the communication frequency needs to be increased to ensure real-time performance. If the device has strong computing performance, it can process communication data more efficiently and appropriately increase the communication frequency to support complex tasks; on the contrary, when the performance is weak, the communication frequency needs to be appropriately reduced to avoid device overload.

[0099] When the data processing location is an ocean platform or unmanned boat, the communication frequency needs to be dynamically adjusted according to the relative size of the task complexity index and the computing performance index. If the computing performance index is high (the equipment has strong computing adaptability) or the task complexity index is low (the task is simple), the frequency adjustment range is small. If the computing performance index is low or the task complexity index is high, the frequency adjustment range is large to ensure that the communication resources can meet the needs of complex tasks.

[0100] When the data processing location is in the cloud, the adjustment of the communication frequency mainly depends on the ratio of the task complexity index and the task quality standard value. High task complexity index: The communication frequency needs to be increased to meet the complexity of the task.

[0101] By adjusting the communication frequencies between the ocean platform and the unmanned boat, as well as with the cloud, targeted optimization can be performed according to the characteristics of the data processing location: Ocean platform and unmanned boat: pay more attention to the adaptability of computing performance and the complexity of tasks, ensuring efficient resource allocation during local processing. Cloud: Considering the powerful computing power of the cloud, pay more attention to the requirements of the task on communication quality and task complexity, ensure high-quality remote data processing, and the adjustment of communication frequency can not only meet the task requirements, but also save bandwidth resources and reduce energy consumption by reducing unnecessary communication frequencies, thus optimizing the overall operation efficiency of the system.

[0102] The above formulas are all dimensionless and numerical calculations. The formula is a formula for the most recent real situation obtained by collecting a large amount of data and performing software simulation. The preset parameters in the formula are set by technicians in this field according to actual conditions.

[0103] It should be understood that in the various embodiments of the present application, the size of the serial numbers of the above-mentioned processes does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present application.

[0104] Those of ordinary skill in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

[0105] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0106] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art who is familiar with the present technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

Claims

1. A collaborative navigation data management method between an offshore platform and an unmanned boat, characterized in that: The following steps are involved: The marine environment data is collected in real time to obtain a marine environment data set, and the marine platform operation data, the unmanned boat operation data and the collaborative navigation task data are obtained to obtain a marine platform operation data set, an unmanned boat operation data set and a collaborative navigation task data set; Based on the marine platform operation data set, the unmanned boat operation data set and the collaborative navigation task data set, a collaborative analysis operation is performed to generate a computing performance index and a task complexity index, and then the computing performance index and the task complexity index are fuzzy reasoned together, and the location of the marine environment data set for data processing is determined according to the result of the fuzzy reasoning; At the position corresponding to the result of fuzzy reasoning, data processing is performed on the marine environment data set to obtain a processed data set, and navigation path planning is performed according to the processed data set; Based on the computing performance index and the task complexity index, the communication frequency of the communication path is optimized and adjusted; The logic for obtaining the calculation performance index is: Obtain the computing performance of the ocean platform and mark it as Pplatform, the computing performance of the unmanned boat and mark it as Pdrone, the computing load occupancy rate of the ocean platform and mark it as Lplatform, the computing load occupancy rate of the unmanned boat and mark it as Ldrone, and substitute them into the following calculation formula: ; represents the computing performance index, represents the dynamic load adjustment factor, represents the resource balancing factor; The resource balancing factor is calculated as follows: ; Dynamic load adjustment factor Obtained through the following calculation: ; A constant set to prevent the denominator from being zero; The logic for obtaining the task complexity index is: The entire navigation area is divided into multiple sub-areas, and the historical data and real-time data of the ocean current speed in each sub-area are obtained. For each sub-area, the absolute value of the difference between the current ocean current speed and the ocean current speed value in the previous time window is calculated first, and then the absolute value is divided by the ocean current speed value in the previous time window to obtain the change amplitude of the ocean current speed. Then, the change amplitude values ​​of all sub-areas are weighted averaged to obtain the environmental complexity coefficient of the entire navigation area. When calculating the weighted average, the weight coefficient of each sub-area is defined as the reciprocal of the square of the Euclidean distance between the coordinates of the center point of the sub-area and the current coordinates of the unmanned boat; Divide the number of obstacles in the entire navigation area by the entire navigation area to obtain the obstacle complexity coefficient; Get the straight-line distance between the current coordinates of the unmanned boat and the destination coordinates and mark it as LL, the number of waypoints that need to be passed in the entire navigation area and mark it as TJ, the preset path planning tolerance update delay time value and mark it as YC, and then calculate: ; Represents the task tolerance coefficient, w1 and w2 are both preset proportional coefficients; Finally, the task tolerance coefficient , obstacle complexity coefficient, and environment complexity coefficient are substituted into the pre-trained neural network model, and the output result of the neural network model is the task complexity index; Based on the computing performance index and the task complexity index, optimizing the communication frequency of the communication path means: If the location for processing the marine environment dataset is an offshore platform or unmanned boat, make the following adjustments: ; The task complexity index is is the basic communication frequency, is the preset adjustment coefficient 1, Indicates the communication frequency of the communication link between the ocean platform and the unmanned boat after adjustment; If the location for processing the marine environment dataset is the cloud, make the following adjustments: ; is the preset task quality standard value, is the preset adjustment factor 2, It indicates the communication link between the marine platform and the cloud, and the communication link between the unmanned boat and the cloud after adjustment.

2. The collaborative navigation data management method between an ocean platform and an unmanned boat according to claim 1, characterized in that: Data processing of the marine environment data set refers to intelligently classifying and integrating the various marine environment data in the marine environment data set according to the preset processing strategy.

3. The collaborative navigation data management method between an ocean platform and an unmanned boat according to claim 2, characterized in that: When performing navigation path planning, select one of the following planning algorithms: a graph-based path planning algorithm, an optimization-based path planning algorithm, and a deep learning-based path planning algorithm.

4. The collaborative navigation data management method between an ocean platform and an unmanned boat according to claim 3, characterized in that: The logic of fuzzy reasoning is: The computing performance index and the task complexity index are taken as input variables, and the location for data processing is taken as the output variable. The input variables are fuzzified and the values ​​of the input variables are converted into fuzzy sets. The output variables are fuzzified and the output variables are converted into fuzzy sets. Fuzzy rules are formulated to describe the degree of data processing adaptation under different data type combinations. The fuzzified input variables are inferred through fuzzy rules to obtain the location for data processing of the marine environment dataset. The location for data processing of the marine environment dataset is one of the marine platform, unmanned boat, and cloud.

Images