Marine emergency information transmission method and system based on satellite communication

Abstract

The invention provides a marine emergency information transmission method and system based on satellite communication. The method comprises the following steps: constructing a temporary special satellite communication link, and performing parameter optimization on the temporary special satellite communication link to obtain an optimized communication link; performing three-dimensional reconstruction on the field environment, analyzing the severity description of the illness state, predicting the illness state development, and generating an illness state prediction result; the distribution strategy of the special medical supplies is optimized by considering factors of weather changes and ocean current directions, and an optimal distribution path plan is obtained; performing pattern recognition and trend prediction, generating potential risk early warning, and forming updated medical requirements; and re-evaluating and optimizing the optimal distribution path plan, and generating an updated optimal distribution path plan. According to the technical scheme provided by the invention, the efficiency and accuracy of maritime first-aid response are improved, and medical supplies can be timely and accurately sent to a first-aid site.

Description

Technical Field

[0001] The embodiments of the present application relate to the technical field of maritime emergency information transmission, and in particular, to a maritime emergency information transmission method and system based on satellite communication. Background Art

[0002] In maritime emergency rescue scenarios, timely and accurate medical response is crucial. Due to the complex and changeable marine environment, traditional land-based emergency rescue methods are difficult to apply directly to maritime rescue. Specifically, maritime emergency rescue needs to ensure stable, low-latency high-definition audio and video calls and data sharing between medical experts and on-site rescue personnel. Accurate on-site environment reconstruction and disease analysis tools must be provided to assist medical experts in making accurate diagnoses and treatment recommendations. In view of factors such as weather changes and ocean current direction, the distribution routes of special medical supplies need to be continuously optimized to ensure that the supplies can be delivered to the emergency site in a timely and accurate manner.

[0003] At present, the transmission of emergency information at sea mainly relies on conventional satellite communication systems and limited remote diagnostic tools. These solutions usually include basic voice and low-bandwidth data transmission, but lack optimization for emergency scenarios. For example, video conferencing software can provide basic audio and video communication, but it is insufficient in image quality and real-time performance. Material distribution plans based on preset conditions cannot be dynamically adjusted to adapt to real-time changing environmental factors.

[0004] However, existing solutions have shown obvious limitations when dealing with complex maritime emergency rescue missions. Standard satellite communication links have not been optimized for parameters and are easily interfered by external factors such as weather, resulting in communication delays or interruptions, affecting the efficiency and accuracy of remote diagnosis. Existing remote guidance tools lack the support of augmented reality technology and computer vision algorithms, and cannot provide comprehensive and intuitive on-site three-dimensional reconstruction, which limits medical experts' understanding and decision-making capabilities of on-site conditions. Static distribution planning fails to fully consider factors such as weather changes and ocean current direction, making the material distribution route inflexible, which may cause material delays or failure to arrive at the emergency site on time, thus affecting the treatment effect.

[0005] In summary, the existing maritime emergency information transmission scheme has significant deficiencies in communication quality, remote diagnosis accuracy and material distribution flexibility, and a more efficient, accurate and flexible method is urgently needed to meet the actual needs of maritime emergency. The present invention aims to comprehensively improve the information transmission and response capabilities of maritime emergency by constructing a temporary dedicated satellite communication link, combining augmented reality technology and computer vision algorithms for three-dimensional reconstruction, applying machine learning models to predict the progression of the disease, and optimizing the material distribution path through reinforcement learning algorithms. Summary of the invention

[0006] The embodiments of the present application provide a method and system for transmitting maritime emergency information based on satellite communication, so as to solve the problem of low efficiency and accuracy of maritime emergency response in the prior art.

[0007] In a first aspect, an embodiment of the present application provides a method for transmitting emergency information at sea based on satellite communication, comprising:

[0008] Building a temporary dedicated satellite communication link according to the emergency location information and the preliminary medical condition description in the received emergency medical distress signal, and optimizing the parameters of the temporary dedicated satellite communication link to obtain an optimized communication link;

[0009] The optimized communication link is used to perform three-dimensional reconstruction of the on-site environment in combination with augmented reality technology and computer vision algorithms, so as to realize high-definition audio and video calls and real-time data sharing between medical experts and on-site rescue personnel, and a machine learning model is used to analyze the description of the severity of the disease, predict the progression of the disease, and generate a disease prediction result;

[0010] Based on the disease prediction results, the distribution strategy of special medical supplies is optimized by taking into account factors such as weather changes and ocean current direction through a reinforcement learning algorithm to obtain the optimal distribution route planning;

[0011] According to the optimized communication link, vital sign monitoring data and specific key environmental indicators are collected and transmitted back to the medical command center in real time, and deep learning algorithms are applied to perform pattern recognition and trend prediction to generate potential risk warnings and form updated medical needs. The specific key environmental indicators include ambient temperature, humidity, air pressure, wind speed, rainfall and ocean current speed;

[0012] Based on the potential risk warning and updated medical needs, the optimal distribution route plan is re-evaluated and optimized, and an updated optimal distribution route plan is generated to ensure that supplies can be delivered to the emergency site in a timely and accurate manner.

[0013] Optionally, the optimized communication link is used in combination with augmented reality technology and computer vision algorithms to perform three-dimensional reconstruction of the on-site environment, realize high-definition audio and video calls and real-time data sharing between medical experts and on-site rescue personnel, and apply a machine learning model to analyze the description of the severity of the disease, predict the progression of the disease, and generate a disease prediction result, including:

[0014] By using the optimized communication link, a high-quality communication connection is established between the emergency site and the medical experts, ensuring stable, low-latency high-definition audio and video calls. The optimized communication link supports large-bandwidth data transmission, ensures the quality and efficiency of real-time data sharing, and obtains a high-quality communication connection;

[0015] Based on the high-quality communication connection, combined with augmented reality technology, the emergency scene images and video streams are processed to achieve superimposition of virtual information on the actual scene at the medical expert end to assist remote diagnosis and obtain an augmented reality-assisted remote diagnosis environment;

[0016] Using computer vision algorithms, multi-angle images and videos transmitted from the scene and processed by augmented reality technology are processed to build an accurate three-dimensional model of the emergency scene, and based on the data obtained in the augmented reality-assisted remote diagnosis environment, an accurate three-dimensional reconstruction of the scene is obtained;

[0017] According to the optimized communication link and the accurate on-site three-dimensional reconstruction, the vital signs data and specific key environmental indicators transmitted from various on-site monitoring equipment are shared in real time and displayed in the form of charts or numerical values ​​on the expert end interface, so as to facilitate instant assessment of the patient's status, promote collaborative work, and obtain real-time shared vital signs data, wherein the specific key environmental indicators include ambient temperature, humidity, air pressure, wind speed, rainfall, and ocean current speed;

[0018] Preparing input data required by the machine learning model, organizing and preprocessing the preliminary medical condition description and the vital signs data shared in real time through the communication link, and forming a structured data set suitable for machine learning model analysis based on the real-time shared vital signs data;

[0019] Applying a machine learning model to perform quantitative evaluation and trend prediction on the structured data set suitable for machine learning model analysis, the machine learning model can quantitatively evaluate the severity of the disease and predict the development trend after learning from a large number of cases, and obtain detailed disease analysis results;

[0020] A disease prediction report is generated based on the detailed disease analysis results and the precise on-site three-dimensional reconstruction.

[0021] Optionally, the computer vision algorithm is used to process the multi-angle images and videos transmitted from the scene and processed by augmented reality technology to construct an accurate three-dimensional model of the emergency scene, and based on the data obtained in the augmented reality-assisted remote diagnosis environment, an accurate three-dimensional reconstruction of the scene is obtained, including:

[0022] Using computer vision algorithms, the received multi-angle images and videos are pre-processed to remove noise, correct distortion, and perform color calibration to obtain high-quality input data;

[0023] Based on the high-quality input data, using a feature detection algorithm, identifying and extracting stable feature points to obtain key feature points for subsequent matching;

[0024] Based on the key feature points used for subsequent matching, a feature matching algorithm is used to match the key feature points under different viewing angles, and the matching results are optimized by bundle adjustment to improve the accuracy of three-dimensional reconstruction and generate optimized feature point matching results;

[0025] The positions of the feature points in the optimized feature point matching results in three-dimensional space are calculated by using the triangulation principle to preliminarily form the spatial structure of the emergency scene and obtain a sparse three-dimensional point cloud;

[0026] Applying a multi-view stereo vision algorithm to fill the gaps between the sparse three-dimensional point clouds to generate a dense three-dimensional point cloud;

[0027] A Poisson surface reconstruction algorithm is used to construct a continuous three-dimensional surface from the dense three-dimensional point cloud, and texture information of the original image is mapped onto the surface to obtain an accurate three-dimensional model;

[0028] The precise three-dimensional model is combined with real-time scene information obtained through augmented reality technology to provide medical experts with a comprehensive and intuitive view of the emergency scene, ensure the accuracy and effectiveness of remote diagnosis, and obtain precise three-dimensional reconstruction of the scene.

[0029] Optionally, based on the disease prediction result, the distribution strategy of special medical supplies is optimized by taking into account factors such as weather changes and ocean current direction through a reinforcement learning algorithm to obtain an optimal distribution route planning, including:

[0030] Based on the disease prediction results, the comprehensive environmental factor assessment module is used to obtain and analyze key environmental parameters of weather changes and ocean current directions in real time to obtain a report on environmental conditions that affect delivery route planning;

[0031] Based on the environmental condition report, a reinforcement learning algorithm is used to create a multi-objective optimization problem model including transportation cost, time efficiency, and safety factor to obtain an initial distribution strategy;

[0032] A dynamic adjustment mechanism is introduced into the initial distribution strategy, allowing the distribution strategy to be updated in real time with the latest weather changes and ocean current directions, ensuring that the effectiveness and accuracy of distribution can be maintained even in complex and changing environments, and generating a continuously optimized distribution path plan;

[0033] By utilizing the interactive learning process between the reinforcement learning model and the actual delivery situation, the delivery route planning is automatically adjusted, and the optimal delivery route planning is obtained based on the latest environmental changes and delivery feedback.

[0034] Optionally, based on the environmental condition report, a reinforcement learning algorithm is applied to create a multi-objective optimization problem model including transportation cost, time efficiency, and safety factor to obtain an initial distribution strategy, including:

[0035] Using the environmental condition report, the key environmental parameters of weather changes and ocean current directions are deeply analyzed and processed to obtain the analysis results of factors affecting the distribution route selection and material delivery time;

[0036] Based on the factor analysis results, multiple optimization objectives are defined, including minimizing transportation costs, maximizing time efficiency, and ensuring the highest safety factor, to generate a multi-objective optimization framework;

[0037] Applying a reinforcement learning algorithm, based on the multi-objective optimization framework, constructing a multi-objective optimization problem model capable of handling the trade-off relationship between different optimization objectives, and using the objective definition and weight allocation in the multi-objective optimization framework as the basis for model construction;

[0038] Constructing a simulation environment to simulate different delivery scenarios, providing a training platform for the multi-objective optimization problem model, and applying the multi-objective optimization model to the simulation environment;

[0039] Through a large number of simulation experiments, the reinforcement learning model is allowed to continuously try to select the delivery path in the simulation environment, and the trade-off relationship between the various objectives is adjusted according to the feedback, gradually converging to the optimal strategy combination, and obtaining a preliminary optimized delivery plan;

[0040] Based on the preliminary optimized distribution plan, combined with the latest environmental conditions and actual distribution needs, final adjustments are made to generate an initial distribution strategy.

[0041] Optionally, according to the optimized communication link, vital sign monitoring data and specific key environmental indicators are collected and transmitted back to the medical command center in real time, and deep learning algorithms are applied to perform pattern recognition and trend prediction to generate potential risk warnings and form updated medical needs. The specific key environmental indicators include ambient temperature, humidity, air pressure, wind speed, rainfall and ocean current speed, including:

[0042] By using the optimized communication link, vital sign monitoring data and specific key environmental indicators at the emergency scene are collected and processed to obtain encoded monitoring data, wherein the specific key environmental indicators include ambient temperature, humidity, air pressure, wind speed, rainfall and ocean current speed;

[0043] Based on the encoded monitoring data, the encoded monitoring data is processed in real time through the optimized communication link to obtain a data stream that is transmitted to a medical command center in real time;

[0044] According to the data stream transmitted to the medical command center in real time, applying the deep learning algorithm deployed in the medical command center, analyzing and processing the vital sign monitoring data and specific key environmental indicators to obtain a pattern recognition result, wherein the specific key environmental indicators include ambient temperature, humidity, air pressure, wind speed, rainfall, and ocean current speed;

[0045] Based on the pattern recognition results, a deep learning model is used to predict the development trend of the patient's condition and generate a trend prediction report;

[0046] Based on the trend prediction report, the system automatically generates a potential risk warning for each patient, and obtains warning information including current risk factors and future risk warnings;

[0047] Based on the warning information, the medical command center re-evaluates and adjusts the emergency plan to form updated medical needs.

[0048] Optionally, based on the potential risk warning and the updated medical needs, the optimal distribution path plan is re-evaluated and optimized to generate an updated optimal distribution path plan to ensure that the supplies can be delivered to the emergency site in a timely and accurate manner, including:

[0049] Utilize the potential risk warning to analyze and process the risk factors at the emergency scene and obtain a detailed risk assessment report;

[0050] Based on the risk assessment report and the updated medical needs, reassess the current medical supplies needs and generate an updated medical supplies needs list;

[0051] Based on the latest medical supplies demand list, check whether the existing optimal distribution route planning needs to be adjusted to adapt to new demand changes, and preliminarily determine the aspects that need to be optimized;

[0052] Based on the above-mentioned optimization aspects, the optimal distribution route is replanned taking into account the latest environmental conditions to ensure that the material distribution strategy meets the latest medical needs and risk warnings;

[0053] Through continuous updating and adjustment, an updated optimal distribution route plan is eventually generated to ensure that special medical supplies can be delivered to the emergency site in a timely and accurate manner.

[0054] In a second aspect, an embodiment of the present application provides a maritime emergency information transmission system based on satellite communication, comprising:

[0055] Constructing an optimization module, for constructing a temporary dedicated satellite communication link according to the emergency location information and the preliminary medical condition description in the received emergency medical distress signal, and optimizing the parameters of the temporary dedicated satellite communication link to obtain an optimized communication link;

[0056] A reconstruction and analysis module is used to utilize the optimized communication link, combined with augmented reality technology and computer vision algorithms, to perform three-dimensional reconstruction of the on-site environment, to achieve high-definition audio and video calls and real-time data sharing between medical experts and on-site rescue personnel, and to apply a machine learning model to analyze the description of the severity of the disease, predict the progression of the disease, and generate a disease prediction result;

[0057] An optimization processing module is used to optimize the distribution strategy of special medical supplies based on the disease prediction results, by using a reinforcement learning algorithm and taking into account factors such as weather changes and ocean current direction, to obtain an optimal distribution route plan;

[0058] A collection and formation module is used to collect vital sign monitoring data and specific key environmental indicators according to the optimized communication link, transmit them back to the medical command center in real time, and apply deep learning algorithms to perform pattern recognition and trend prediction, generate potential risk warnings, and form updated medical needs. The specific key environmental indicators include ambient temperature, humidity, air pressure, wind speed, rainfall, and ocean current speed;

[0059] The evaluation and optimization module is used to re-evaluate and optimize the optimal distribution route plan based on the potential risk warning and updated medical needs, generate an updated optimal distribution route plan, and ensure that the supplies can be delivered to the emergency site in a timely and accurate manner.

[0060] In a third aspect, an embodiment of the present application provides a computing device, comprising a processing component and a storage component; the storage component stores one or more computer instructions; the one or more computer instructions are used to be called and executed by the processing component to implement a method for transmitting maritime emergency information based on satellite communication as described in any one of the first aspects.

[0061] In a fourth aspect, an embodiment of the present application provides a computer storage medium storing a computer program. When the computer program is executed by a computer, it implements a method for transmitting maritime emergency information based on satellite communication as described in any one of the first aspects.

[0062] In the embodiment of the present application, a temporary dedicated satellite communication link is constructed according to the emergency location information and preliminary medical condition description in the received emergency medical distress signal, and the parameters of the temporary dedicated satellite communication link are optimized to obtain an optimized communication link; the optimized communication link is used to combine augmented reality technology and computer vision algorithms to perform three-dimensional reconstruction of the on-site environment, realize high-definition audio and video calls and real-time data sharing between medical experts and on-site rescue personnel, and apply a machine learning model to analyze the description of the severity of the disease, predict the development of the disease, and generate a disease prediction result; based on the disease prediction result, a reinforcement learning algorithm is used to consider weather changes, ocean currents, and other factors. Based on the factors of direction, the distribution strategy of special medical supplies is optimized to obtain the optimal distribution route planning; according to the optimized communication link, vital sign monitoring data and specific key environmental indicators are collected and transmitted back to the medical command center in real time, and deep learning algorithms are applied to perform pattern recognition and trend prediction, generate potential risk warnings, and form updated medical needs. The specific key environmental indicators include ambient temperature, humidity, air pressure, wind speed, rainfall and ocean current speed; based on the potential risk warnings and updated medical needs, the optimal distribution route planning is re-evaluated and optimized to generate an updated optimal distribution route planning to ensure that the supplies can be delivered to the emergency site in a timely and accurate manner.

[0063] The technical solution of this application has the following beneficial effects:

[0064] This application ensures stable, low-latency high-definition audio and video calls and real-time data sharing by building a temporary dedicated satellite communication link and optimizing the parameters of the communication link, thereby speeding up the information exchange between medical experts and on-site rescue personnel and shortening the decision-making time. The on-site environment is reconstructed in three dimensions using augmented reality technology and computer vision algorithms, and the description of the severity of the disease is analyzed in combination with a machine learning model to predict the development of the disease and generate disease prediction results. This method not only improves the accuracy of remote diagnosis, but also provides a scientific basis for subsequent treatment. Based on the disease prediction results, the distribution strategy of special medical supplies is optimized by considering factors such as weather changes and ocean current direction through a reinforcement learning algorithm to obtain the optimal distribution path planning. This ensures that the supplies can arrive at the emergency site in the shortest time with the lowest risk, and improves the reliability and timeliness of material distribution. Vital sign monitoring data and specific key environmental indicators are collected, and deep learning algorithms are applied for pattern recognition and trend prediction to generate potential risk warnings. This enables the system to better cope with the complex and changing marine environment, predict possible risks in advance, adjust distribution strategies, and ensure the safety and effectiveness of the emergency process. Based on potential risk warnings and updated medical needs, the optimal distribution route planning is re-evaluated and optimized to generate an updated optimal distribution route planning. This dynamic adjustment mechanism ensures that the distribution route is always the best choice, and can respond quickly even in emergencies to ensure timely and accurate delivery of supplies.

[0065] Furthermore, the embodiment of the present application also utilizes the optimized communication link combined with augmented reality technology and computer vision algorithms to perform three-dimensional reconstruction of the emergency scene, realize high-definition audio and video calls and real-time data sharing between medical experts and on-site rescue personnel, and apply machine learning models to analyze and predict the severity of the disease. Through the reinforcement learning algorithm, factors such as weather changes and ocean current direction are considered, and the distribution strategy of special medical supplies is optimized for multiple objectives to ensure that the supplies can be delivered to the emergency site in a timely and accurate manner. This process not only covers the establishment of high-quality communication connections, accurate three-dimensional reconstruction, real-time vital signs data sharing, and the preparation of structured data sets, but also includes the generation of disease prediction reports based on detailed disease analysis results, and the continuous optimization of distribution route planning through a dynamic adjustment mechanism.

[0066] Through the above methods, not only a stable, low-latency high-definition audio and video communication connection and accurate on-site 3D reconstruction are established, but also the accuracy of remote diagnosis is enhanced, providing medical experts with an intuitive and comprehensive view of the first aid scene. At the same time, the machine learning model is applied to quantitatively assess the condition and predict the development trend, generate detailed condition analysis results, and support more scientific decision-making. In addition, by introducing a reinforcement learning algorithm to create a multi-objective optimization problem model including transportation cost, time efficiency, and safety factor, the optimal planning of the distribution path of special medical supplies is achieved, and the effectiveness and accuracy of distribution can be maintained even in a complex and changeable marine environment. These improvements have jointly improved the overall performance of the system, maximized the safety of patients' lives, and significantly increased the success rate of maritime first aid missions.

[0067] These and other aspects of the present application will become more clearly understood in the description of the following embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0068] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, a brief introduction will be given below to the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0069] Figure 1 A flowchart of a method for transmitting emergency information at sea based on satellite communication provided in an embodiment of the present application;

[0070] Figure 2 A schematic diagram of the structure of a satellite communication-based marine emergency information transmission system provided in an embodiment of the present application;

[0071] Figure 3 A schematic diagram of the structure of a computing device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0072] In order to enable those skilled in the art to better understand the solution of the present application, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application.

[0073] In some of the processes described in the specification and claims of this application and the above-mentioned figures, multiple operations that appear in a specific order are included, but it should be clearly understood that these operations may not be performed in the order in which they appear in this article or may be performed in parallel. In addition, these processes may include more or fewer operations, and these operations may be performed in sequence or in parallel. It should be noted that the descriptions of "first", "second", etc. in this article are used to distinguish different messages, devices, modules, etc., and do not represent the order of precedence, nor do they limit "first" and "second" to be different types.

[0074] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of this application.

[0075] Figure 1 A flowchart of a method for transmitting emergency information at sea based on satellite communication is provided for an embodiment of the present application, such as Figure 1 As shown, the method includes:

[0076] Building a temporary dedicated satellite communication link according to the emergency location information and the preliminary medical condition description in the received emergency medical distress signal, and optimizing the parameters of the temporary dedicated satellite communication link to obtain an optimized communication link;

[0077] According to the emergency location information and preliminary medical condition description in the received emergency medical distress signal, a temporary dedicated satellite communication link is constructed, and the parameters of the temporary dedicated satellite communication link are optimized to obtain an optimized communication link. The emergency location information usually includes data such as longitude and latitude coordinates and depth (if applicable) to accurately locate the location of the emergency scene. The preliminary medical condition description contains the patient's basic symptoms, vital signs and other information, which is crucial for quickly assessing the severity of the disease. A temporary dedicated satellite communication link refers to a communication connection established specifically for the current emergency mission, which can ensure stable, low-latency data transmission. Parameter optimization refers to adjusting the various parameters of the communication link (such as frequency, bandwidth, power, etc.) to adapt to specific marine environments and emergency needs.

[0078] After receiving an emergency medical distress signal, the system first parses the emergency location information and preliminary medical condition description contained therein. Based on this information, the system selects the optimal satellite resources and builds a temporary dedicated satellite communication link. Then, the system dynamically adjusts the link parameters according to real-time environmental conditions (such as weather, ocean currents, etc.) to ensure the stability and efficiency of the communication link. Ultimately, this optimized communication link will provide a solid foundation for subsequent high-definition audio and video calls and real-time data sharing.

[0079] In an actual marine emergency scenario, a cruise ship sent out an emergency medical distress signal, reporting that a passenger had a sudden heart attack. After receiving the signal, the system quickly parsed the precise location of the cruise ship (for example, 40.7128°N, 74.0060°W) and the patient's preliminary medical condition description (such as abnormal heart rate, difficulty breathing). Subsequently, the system automatically selected the most suitable satellite resources and established a temporary dedicated satellite communication link from the cruise ship to the onshore medical command center. Next, the system adjusted the parameters of the communication link according to the weather forecast and ocean current data at the time, ensuring the stability and efficiency of the communication link in bad weather, thereby providing reliable communication guarantee for subsequent remote diagnosis and material distribution.

[0080] The optimized communication link is used to perform three-dimensional reconstruction of the on-site environment in combination with augmented reality technology and computer vision algorithms, so as to realize high-definition audio and video calls and real-time data sharing between medical experts and on-site rescue personnel, and a machine learning model is used to analyze the description of the severity of the disease, predict the progression of the disease, and generate a disease prediction result;

[0081] It involves using optimized communication links, combined with augmented reality technology and computer vision algorithms to reconstruct the emergency scene in three dimensions, realize high-definition audio and video calls and real-time data sharing between medical experts and on-site rescue personnel, and apply machine learning models to analyze the description of the severity of the disease, predict the development of the disease, and generate disease prediction results. Augmented reality technology allows virtual information to be superimposed on the actual scene to assist remote diagnosis; computer vision algorithms construct accurate three-dimensional models by processing multi-angle images and videos; machine learning models are used to quantitatively evaluate the disease and predict development trends.

[0082] Through the optimized communication link, medical experts can have stable high-definition audio and video calls with on-site rescue personnel, and share images and videos of the emergency scene in real time. At the same time, computer vision algorithms process these images and videos to build accurate three-dimensional models to help medical experts understand the on-site situation more intuitively. In addition, machine learning models analyze the collected preliminary medical condition descriptions and real-time monitoring data to generate detailed disease prediction results to support more scientific decision-making.

[0083] Continuing with the cruise ship case above, after establishing a stable communication link, medical experts guided the on-site rescue personnel through high-definition audio and video calls. At the same time, computer vision algorithms processed multi-angle images and videos sent back from the cruise ship to build an accurate three-dimensional model of the interior of the cruise ship and around the patient. Medical experts use augmented reality technology to see the actual scene with virtual information superimposed on it, and better understand the situation on the scene. The machine learning model predicts the possible development trend of the disease based on the patient's vital signs and other key indicators, and generates a detailed disease prediction report, which provides a scientific basis for the next step of treatment.

[0084] Based on the disease prediction results, the distribution strategy of special medical supplies is optimized by taking into account factors such as weather changes and ocean current direction through a reinforcement learning algorithm to obtain the optimal distribution route planning;

[0085] Based on the disease prediction results, the reinforcement learning algorithm takes into account factors such as weather changes and ocean current direction, optimizes the distribution strategy of special medical supplies, and obtains the optimal distribution route planning. The disease prediction results provide detailed information about the development of the patient's condition, and the reinforcement learning algorithm simulates different distribution routes to find the most effective distribution plan to ensure that the supplies can reach the emergency site in the shortest time with the lowest risk.

[0086] Based on the generated disease prediction results, the system uses the comprehensive environmental factor assessment module to obtain key environmental parameters such as real-time weather changes and ocean current direction, and generates an environmental condition report. Then, a reinforcement learning algorithm is applied to create a multi-objective optimization problem model, taking into account factors such as transportation cost, time efficiency and safety factor, and generating an initial distribution strategy. With the continuous update of the latest weather changes and ocean current direction data, the distribution strategy is adjusted accordingly to ensure that the effectiveness and accuracy of distribution can be maintained even in complex and changing environments, and ultimately generate the optimal distribution path planning.

[0087] In the case of a cruise ship patient with heart disease, based on the disease prediction report, the system determined that a defibrillator and drugs needed to be delivered to the cruise ship immediately. The system obtains real-time weather change and current direction data through the comprehensive environmental factor evaluation module and generates an environmental condition report. Next, a reinforcement learning algorithm is applied to create a multi-objective optimization problem model that includes transportation cost, time efficiency, and safety factor to generate an initial delivery strategy. With the continuous update of weather and current data, the delivery route has been optimized and adjusted many times to ensure that the defibrillator and drugs can be safely delivered to the cruise ship in the shortest time, buying precious time for treating patients.

[0088] Based on the optimized communication link, vital sign monitoring data and specific key environmental indicators are collected, transmitted back to the medical command center in real time, and pattern recognition and trend prediction are carried out using deep learning algorithms to generate potential risk warnings and form updated medical needs. The specific key environmental indicators include environmental temperature, humidity, air pressure, wind speed, rainfall, and sea current speed.

[0089] Based on the optimized communication link, vital sign monitoring data and specific key environmental indicators are collected, transmitted back to the medical command center in real time, and pattern recognition and trend prediction are carried out using deep learning algorithms to generate potential risk warnings and form updated medical needs. The vital sign monitoring data includes heart rate, blood pressure, blood oxygen saturation, etc., which reflect the health status of the patient. The specific key environmental indicators cover environmental temperature, humidity, air pressure, wind speed, rainfall, and sea current speed, and these data affect the safety of material distribution and rescue operations.

[0090] Through the optimized communication link, the system collects in real time the vital sign data and specific key environmental indicators transmitted by various monitoring devices at the first-aid scene. These data are transmitted back to the medical command center in real time, and pattern recognition and trend prediction are carried out through deep learning algorithms to generate potential risk warnings. Based on these warnings, the system will re-evaluate and update the medical needs to ensure that all necessary medical supplies and services can be in place in a timely manner.

[0091] In the case of a heart disease patient on a cruise ship, the system continuously collects the vital sign data of the patient, such as heart rate, blood pressure, blood oxygen saturation, etc., and key environmental indicators such as environmental temperature, humidity, air pressure, wind speed, rainfall, and sea current speed, through the optimized communication link. These data are transmitted back to the medical command center in real time and analyzed through deep learning algorithms, and it is found that the patient may be at risk of acute myocardial infarction. Based on this potential risk warning, the system immediately updates the medical needs, increases the demand for first-aid drugs, and notifies relevant units to prepare further rescue measures to ensure that the patient receives the most timely and appropriate treatment.

[0092] Based on the potential risk warning and the updated medical needs, re-evaluate and optimize the optimal delivery route planning, and generate an updated optimal delivery route planning to ensure that materials can be delivered to the first-aid scene in a timely and accurate manner.

[0093] Based on the potential risk warning and the updated medical needs, re-evaluate and optimize the optimal delivery route planning to ensure that materials can be delivered to the first-aid scene in a timely and accurate manner. The potential risk warning provides information on possible dangerous situations in the future, while the updated medical needs clarify the new requirements for materials and services. By re-evaluating and optimizing the delivery route, the system ensures the flexibility and reliability of material distribution and improves the overall efficiency of the first-aid response.

[0094] The system re-evaluates the existing optimal distribution route plan based on the generated potential risk warnings and updated medical needs. Taking into account new risk factors and medical needs, the system once again applies reinforcement learning algorithms to optimize the distribution route to ensure that the supplies can reach the emergency site safely in the shortest time. Through continuous dynamic adjustments, the system ensures that the distribution route is always the best choice and can respond quickly even in emergencies.

[0095] In the case of a cruise ship heart patient, the system discovered the patient's risk of acute myocardial infarction based on potential risk warnings, updated medical needs, and increased the demand for emergency medicines. Based on this new information, the system re-evaluated the original optimal delivery route planning, taking into account the latest weather changes and ocean current direction, and again applied the reinforcement learning algorithm to optimize the delivery route. In the end, the defibrillator and the newly added emergency medicines were safely delivered to the cruise ship by drone in the shortest time, winning precious treatment time for the patient and successfully saving the patient's life.

[0096] The present invention significantly improves the speed and accuracy of emergency response at sea, ensuring that medical supplies can be delivered to the emergency site in a timely and accurate manner. Specifically, a temporary dedicated satellite communication link is constructed and optimized to enhance the stability and efficiency of communication; three-dimensional reconstruction is performed by combining augmented reality technology and computer vision algorithms to improve the accuracy of remote diagnosis; machine learning models are applied to predict the progression of the disease to support more scientific decision-making; the distribution strategy of special medical supplies is optimized through reinforcement learning algorithms to ensure the effectiveness and safety of material distribution; finally, comprehensive monitoring and dynamic adjustment of the emergency process are achieved through real-time collection and analysis of vital sign monitoring data and specific key environmental indicators. These improvements together improve the overall performance of the system, maximize the safety of patients' lives, and significantly enhance the success rate of emergency missions at sea.

[0097] The following is an introduction to the process of the maritime emergency information transmission method based on satellite communication. The entire process from the receipt of the emergency medical distress signal to the final optimization of material distribution is as follows:

[0098] First, when an emergency occurs at sea, a satellite communication terminal (such as terminals of Inmarsat, Iridium, etc.) pre-deployed on a ship or facility can be used to send a distress signal using a global satellite network. The distress signal is forwarded to the nearest ground station (Earth Stat ion) via satellite, which then passes the information to the Rescue Coordination Center (RCC) or medical service providers.

[0099] Secondly, according to the location of the distress call and the required bandwidth requirements, resources are dynamically allocated from the satellite operator to build a temporary dedicated satellite communication link. In view of the characteristics of the marine environment (such as high humidity, strong electromagnetic interference, etc.), the satellite link is optimized in terms of modulation and demodulation schemes, coding strategies, etc. to ensure link stability and data transmission rate.

[0100] Furthermore, through the optimized satellite link, low-latency coding technology is used to achieve high-definition audio and video calls, and good call quality can be maintained even under weak signal conditions. With the two-way communication capability of the satellite link, real-time data sharing is achieved, including patient vital signs data, on-site images, and three-dimensional environmental models.

[0101] Next, ensure that all data involving patient privacy is transmitted through an encrypted tunnel and directly reaches the terminal device of the remote medical expert via a satellite link. The disease prediction model may run on a cloud server, uploading and downloading necessary computing resources and results via a satellite link, reducing the local computing burden.

[0102] Next, the precise location services provided by satellite navigation systems (such as GPS and GLONASS) and weather forecast information provided by meteorological satellites are combined to select the optimal route for material distribution. If there are sudden weather changes or other influencing factors, the distribution instructions can be updated instantly through satellite links to ensure that the materials are delivered on time.

[0103] Then, using the persistent connection provided by the satellite link, vital sign monitoring data and key environmental indicators are continuously collected to ensure the continuity of the information flow. These data are transmitted back to the medical command center via the satellite link, and the deep learning algorithm in the center quickly processes the massive data and generates early warning information.

[0104] Finally, the intelligent dispatch platform based on satellite communications can receive information from multiple sources in real time, including the latest risk warnings and changes in medical needs, so as to make the most reasonable path planning decisions. After each incident, the lessons learned through the satellite link will be used to improve algorithms and technical means to improve the response speed and service quality in similar incidents in the future.

[0105] In this way, it is possible to more clearly demonstrate the key role that satellite communications plays in the entire maritime emergency information transmission process and how satellite communications specifically supports the effective implementation of each step.

[0106] In order to solve the quality problem of high-definition audio and video calls and real-time data sharing, in some embodiments, the optimized communication link is used to combine augmented reality technology and computer vision algorithms to perform three-dimensional reconstruction of the on-site environment, realize high-definition audio and video calls and real-time data sharing between medical experts and on-site rescue personnel, and apply machine learning models to analyze the description of the severity of the disease, predict the development of the disease, and generate disease prediction results, including:

[0107] The optimized communication link is used to establish a high-quality communication connection between the emergency scene and the medical experts, ensuring stable, low-latency high-definition audio and video calls. The optimized communication link supports large-bandwidth data transmission, ensures the quality and efficiency of real-time data sharing, and obtains a high-quality communication connection; based on the high-quality communication connection, combined with augmented reality technology, the emergency scene images and video streams are processed to achieve superimposition of virtual information on the actual scene at the medical expert end, assist remote diagnosis, and obtain an augmented reality-assisted remote diagnosis environment; computer vision algorithms are used to process multi-angle images and videos transmitted back from the scene and processed by augmented reality technology to construct an accurate three-dimensional model of the emergency scene, and based on the data obtained in the augmented reality-assisted remote diagnosis environment, an accurate three-dimensional reconstruction of the scene is obtained; according to the optimized communication link and the accurate three-dimensional reconstruction of the scene, the vital signs data and Specific key environmental indicators are shared in real time and displayed in the form of charts or numerical values ​​on the expert end interface to facilitate instant assessment of patient status, promote collaborative work, and obtain real-time shared vital signs data, wherein the specific key environmental indicators include ambient temperature, humidity, air pressure, wind speed, rainfall, and current speed; prepare input data required for the machine learning model, organize and preprocess the preliminary medical condition description and the vital signs data shared in real time through the communication link, and form a structured data set suitable for machine learning model analysis based on the real-time shared vital signs data; apply the machine learning model to quantitatively evaluate and predict trends of the structured data set suitable for machine learning model analysis; after learning from a large number of cases, the machine learning model can quantitatively evaluate the severity of the disease and predict the development trend to obtain detailed disease analysis results; generate a disease prediction report based on the detailed disease analysis results and the precise on-site three-dimensional reconstruction.

[0108] In this embodiment, the high-quality communication connection includes ensuring the establishment of a stable, low-latency high-definition audio and video call between the emergency site and the medical expert through an optimized satellite communication link. The communication link supports large-bandwidth data transmission, ensuring the quality and efficiency of real-time data sharing, thereby obtaining a high-quality communication connection. This connection not only provides clear audio and video streams, but also supports the rapid transmission of high-resolution images and a large amount of monitoring data. The augmented reality-assisted remote diagnosis environment is based on a high-quality communication connection, combined with augmented reality technology, to process the emergency site images and video streams, so as to realize the superposition of virtual information on the actual scene at the medical expert end, and assist remote diagnosis. Augmented reality technology can display the patient's key vital signs data, disease prediction and other information in the form of virtual tags on the video screen, helping medical experts to understand the on-site situation more intuitively. Accurate on-site three-dimensional reconstruction uses computer vision algorithms to process multi-angle images and videos transmitted back from the scene and processed by augmented reality technology to construct an accurate three-dimensional model of the emergency site. These data come from multiple cameras and sensors, and after algorithm processing, they form a complete three-dimensional spatial structure, providing medical experts with a comprehensive and realistic view of the emergency site. The real-time shared vital signs data is based on the optimized communication link and accurate on-site 3D reconstruction, and collects vital signs data and specific key environmental indicators transmitted from various monitoring devices on site, and displays them in the form of charts or numerical values ​​on the expert end interface. These data include vital signs such as heart rate, blood pressure, and blood oxygen saturation, as well as environmental parameters such as ambient temperature, humidity, air pressure, wind speed, rainfall, and ocean current speed, which facilitate instant assessment of patient status and promote collaborative work. The structured data set suitable for machine learning model analysis is the input data required to prepare the machine learning model. The preliminary medical condition description and the vital signs data shared in real time through the communication link are sorted and preprocessed to form a structured data set suitable for machine learning model analysis. These data are preprocessed by cleaning, normalization and other steps to ensure that their format is consistent and easy to analyze. The detailed disease analysis results are the quantitative evaluation and trend prediction of the structured data set suitable for machine learning model analysis by applying the machine learning model. After learning from a large number of cases, the model can quantitatively evaluate the severity of the disease and predict possible development trends, and finally generate detailed disease analysis results. These results provide a scientific basis for subsequent treatment decisions. The disease prediction report is generated based on detailed disease analysis results and accurate on-site 3D reconstruction. This report not only includes the possibility of disease deterioration, but also includes symptom changes that require special attention, providing comprehensive reference information for medical experts.

[0109] In the embodiment of the present application, firstly, the quality of high-definition audio and video calls between the first aid site and the medical experts is ensured through the optimized communication link, large bandwidth data transmission is supported, and the quality and efficiency of real-time data sharing are guaranteed. Secondly, the augmented reality technology is combined to process the first aid site images and video streams, and the virtual information is superimposed on the actual scene at the medical expert end to assist remote diagnosis. Then, the computer vision algorithm is used to process the multi-angle images and videos transmitted back from the scene, and an accurate three-dimensional model of the first aid site is constructed to provide a comprehensive real view. Further, according to the optimized communication link and three-dimensional reconstruction, vital signs data and specific key environmental indicators are collected and shared in real time, and displayed in the form of charts or numerical values, so as to facilitate the immediate assessment of the patient's status. Furthermore, the preliminary medical condition description and the real-time shared vital signs data are sorted and preprocessed to form a structured data set suitable for machine learning model analysis. Finally, the structured data set is quantitatively evaluated and trend predicted to generate detailed disease analysis results. Based on the detailed disease analysis results and three-dimensional reconstruction, a disease prediction report is generated to provide decision support for medical experts.

[0110] Here is a specific example:

[0111] Suppose in a marine emergency scenario, a cruise ship sends an emergency medical distress signal, reporting a passenger suffering a heart attack. After receiving the signal, the system quickly establishes a high-quality satellite communication link, ensuring the quality of high-definition audio and video calls between the emergency scene and the onshore medical command center. Medical experts use augmented reality technology to see virtual information superimposed on the patient's surroundings, such as the patient's heart rate, respiratory rate and other key vital signs displayed in the form of virtual labels on the video screen, helping them better understand the situation on the scene.

[0112] At the same time, computer vision algorithms process multi-angle images and videos sent back from the cruise ship to build an accurate three-dimensional model of the interior of the cruise ship and around the patient. This allows medical experts to more intuitively understand the site layout and patient location through a three-dimensional view, improving the accuracy of remote diagnosis. The system collects and shares patients' vital signs data (such as electrocardiogram, blood oxygen saturation) and specific key environmental indicators (such as ambient temperature, humidity, air pressure, wind speed, rainfall, and ocean current speed) in real time, and displays them in the form of charts or numerical values ​​on the expert interface, making it convenient for medical experts to instantly assess the patient's status.

[0113] Next, the system prepared a structured data set suitable for machine learning model analysis, sorted and preprocessed the initial medical condition description and real-time shared vital signs data. Then, the machine learning model was applied to quantitatively evaluate and predict trends in this data, generating detailed disease analysis results, including the possibility of disease deterioration and symptom changes that require special attention. Based on these analysis results, the system generated a disease prediction report, providing comprehensive reference information for medical experts and supporting more scientific treatment decisions.

[0114] Through these measures, the system not only improves the accuracy and real-time performance of remote diagnosis, but also provides medical experts with a comprehensive first aid scene view and detailed condition analysis, significantly enhancing the success rate of maritime first aid missions.

[0115] In order to solve the accuracy and real-time problems of three-dimensional reconstruction of the emergency scene, in some embodiments, the computer vision algorithm is used to process the multi-angle images and videos transmitted back from the scene and processed by augmented reality technology to construct an accurate three-dimensional model of the emergency scene, and based on the data obtained in the augmented reality-assisted remote diagnosis environment, an accurate three-dimensional reconstruction of the scene is obtained, including:

[0116] Using computer vision algorithms, the received multi-angle images and videos are preprocessed to remove noise, correct distortion, and perform color calibration to obtain high-quality input data; based on the high-quality input data, a feature detection algorithm is used to identify and extract stable feature points to obtain key feature points for subsequent matching; based on the key feature points for subsequent matching, a feature matching algorithm is used to match the key feature points under different viewing angles, and the matching results are optimized through bundle adjustment to improve the accuracy of three-dimensional reconstruction and generate optimized feature point matching results; the optimized feature point matching results are calculated using the triangulation principle The positions of the feature points in three-dimensional space preliminarily form the spatial structure of the emergency scene and obtain a sparse three-dimensional point cloud; a multi-view stereo vision algorithm is used to fill the gaps between the sparse three-dimensional point cloud to generate a dense three-dimensional point cloud; a Poisson surface reconstruction algorithm is used to construct a continuous three-dimensional surface from the dense three-dimensional point cloud, and the texture information of the original image is mapped onto the surface to obtain an accurate three-dimensional model; the accurate three-dimensional model is combined with real-time scene information obtained through augmented reality technology to provide medical experts with a comprehensive and intuitive view of the emergency scene, ensure the accuracy and effectiveness of remote diagnosis, and obtain accurate three-dimensional reconstruction of the scene.

[0117] In this embodiment, the high-quality input data includes preprocessing the received multi-angle images and videos, removing noise, correcting distortion, and performing color calibration. These preprocessing steps ensure the quality of the input data and improve the accuracy of subsequent feature point detection and matching. The key feature points are stable feature points in the image that are identified and extracted using a feature detection algorithm. These feature points are local areas with uniqueness in the image and are used for subsequent feature matching and three-dimensional reconstruction. The optimized feature point matching result is to match the key feature points under different viewing angles through the feature matching algorithm, and optimize the matching results through the bundle adjustment to improve the accuracy of three-dimensional reconstruction. The optimized matching results provide more accurate spatial position information. The sparse three-dimensional point cloud uses the triangulation principle to calculate the position of the successfully matched feature points in three-dimensional space, and preliminarily forms the spatial structure of the emergency scene. The point cloud generated at this stage is relatively sparse, but it can provide a basic spatial framework. The dense three-dimensional point cloud is to fill the gaps between the sparse three-dimensional point clouds using a multi-view stereo vision algorithm to generate a more detailed dense three-dimensional point cloud. This step makes the model more complete and realistic, and enhances the effect of three-dimensional reconstruction. The precise 3D model uses the Poisson surface reconstruction algorithm to construct a continuous 3D surface from a dense 3D point cloud and maps the texture information of the original image onto the surface. The resulting precise 3D model not only has geometric accuracy, but also retains the color and texture details of the original scene. The augmented reality-assisted remote diagnosis environment combines the precise 3D model with real-time scene information obtained through augmented reality technology to provide medical experts with a comprehensive and intuitive view of the emergency scene, ensuring the accuracy and effectiveness of remote diagnosis.

[0118] In the embodiment of the present application, first, the received multi-angle images and videos are preprocessed to remove noise, correct distortion, and perform color calibration to ensure high-quality input data. Secondly, based on the high-quality input data, a feature detection algorithm is used to identify and extract stable feature points to prepare key feature points for subsequent matching. Next, a feature matching algorithm is used to match key feature points under different viewing angles, and the matching results are optimized by bundle adjustment to improve the accuracy of three-dimensional reconstruction. Then, the position of the feature points in the optimized feature point matching results in three-dimensional space is calculated using the triangulation principle to preliminarily form the spatial structure of the emergency scene and obtain a sparse three-dimensional point cloud. Further, a multi-view stereo vision algorithm is applied to fill the gaps between the sparse three-dimensional point clouds to generate a more detailed dense three-dimensional point cloud, making the model more complete and realistic. Furthermore, a Poisson surface reconstruction algorithm is used to construct a continuous three-dimensional surface from a dense three-dimensional point cloud, and the texture information of the original image is mapped onto the surface to obtain an accurate three-dimensional model. Finally, the accurate three-dimensional model is combined with real-time scene information obtained through augmented reality technology to provide medical experts with a comprehensive and intuitive view of the emergency scene, ensuring the accuracy and effectiveness of remote diagnosis.

[0119] Here is a specific example:

[0120] As another example, consider a maritime emergency scenario where a cruise ship sends an emergency medical distress signal reporting a passenger suffering a heart attack. After receiving the signal, the system quickly establishes a high-quality satellite communication link and begins collecting multi-angle images and video streams of the emergency scene. In order to build an accurate 3D model, the system first pre-processes the received images and videos to remove noise from the images, correct lens distortion, and perform color calibration to ensure the quality of the input data.

[0121] Next, the system uses the SI FT feature detection algorithm to identify and extract stable feature points in the image. These feature points are unique local areas in the image and are used for subsequent feature matching. Through the feature matching algorithm, the system matches feature points from different perspectives and optimizes the matching results through bundle adjustment, improving the accuracy of 3D reconstruction. Subsequently, the system uses the principle of triangulation to calculate the position of the successfully matched feature points in 3D space, preliminarily forming the spatial structure of the emergency scene and obtaining a sparse 3D point cloud.

[0122] To make the model more complete and realistic, the system applied a multi-view stereo vision algorithm to fill the gaps between sparse 3D point clouds and generate dense 3D point clouds. Finally, a Poisson surface reconstruction algorithm was used to construct a continuous 3D surface from the dense 3D point cloud, and the texture information of the original image was mapped onto the surface to obtain an accurate 3D model. The model not only has geometric accuracy, but also retains the color and texture details of the original scene.

[0123] The system combines precise 3D models with real-time scene information obtained through augmented reality technology, providing medical experts with a comprehensive and intuitive view of the emergency scene. Medical experts can see virtual information superimposed on the patient's surroundings through augmented reality devices, such as the patient's heart rate, respiratory rate and other key vital signs displayed in the form of virtual labels on the video screen, helping them better understand the on-site situation and significantly improving the accuracy and effectiveness of remote diagnosis.

[0124] Through these measures, the system not only improves the accuracy and real-time performance of the three-dimensional reconstruction of the emergency scene, but also provides medical experts with a comprehensive view of the emergency scene and detailed condition analysis, significantly enhancing the success rate of maritime emergency missions.

[0125] In order to solve the problem of effectiveness and accuracy of special medical supplies distribution path planning in a complex and changeable environment, in some embodiments, based on the disease prediction results, the distribution strategy of special medical supplies is optimized by taking into account factors such as weather changes and ocean current direction through a reinforcement learning algorithm to obtain the optimal distribution path planning, including:

[0126] According to the disease prediction results, the comprehensive environmental factor assessment module is used to obtain and analyze the key environmental parameters of weather changes and ocean current directions in real time, so as to obtain an environmental condition report that affects the distribution route planning; based on the environmental condition report, a reinforcement learning algorithm is used to create a multi-objective optimization problem model including transportation cost, time efficiency, and safety factor, so as to obtain an initial distribution strategy; a dynamic adjustment mechanism is introduced into the initial distribution strategy, so as to allow the distribution strategy to be updated in real time with the latest weather changes and ocean current directions, so as to ensure that the effectiveness and accuracy of distribution can be maintained even in complex and changeable environments, and to generate continuously optimized distribution route plans; the distribution route planning is automatically adjusted by utilizing the interactive learning process between the reinforcement learning model and the actual distribution situation, so as to obtain the optimal distribution route planning based on the latest environmental changes and distribution feedback.

[0127] In this embodiment, the environmental condition report includes the use of a comprehensive environmental factor assessment module to obtain and analyze key environmental parameters (such as wind speed, temperature, humidity, air pressure, rainfall, current speed, etc.) of weather changes and current directions in real time. These data are used to generate environmental condition reports that affect distribution route planning to ensure that distribution route planning can adapt to changes in the actual environment. The multi-objective optimization problem model is based on the environmental condition report and uses a reinforcement learning algorithm to create a multi-objective optimization problem model including transportation cost, time efficiency, and safety factor. This model not only takes into account the cost-effectiveness of distribution, but also takes into account time and safety, ensuring that the best distribution strategy is found under multiple constraints. The initial distribution strategy is a preliminary distribution plan generated by the multi-objective optimization problem model as the basis for subsequent dynamic adjustment. This strategy takes into account current environmental conditions and medical needs and provides an initial distribution route planning. The dynamic adjustment mechanism is to introduce a dynamic adjustment mechanism that allows the distribution strategy to be updated in real time with the latest weather changes and current directions. This mechanism ensures that the effectiveness and accuracy of distribution can be maintained even in a complex and changing environment, and generates a continuously optimized distribution route plan. The interactive learning process is an interactive learning process that uses a reinforcement learning model and actual distribution conditions to automatically adjust the distribution route planning. The system continuously optimizes the delivery route based on the latest environmental changes and delivery feedback, and ultimately obtains the optimal delivery route planning. This interactive learning process enables the system to continuously improve and adapt to new situations in actual operation.

[0128] In the embodiment of the present application, first, according to the disease prediction results, the comprehensive environmental factor assessment module is used to obtain and analyze the key environmental parameters of weather changes and ocean current directions in real time, and generate an environmental condition report that affects the distribution route planning. Secondly, based on the environmental condition report, a reinforcement learning algorithm is applied to create a multi-objective optimization problem model including transportation cost, time efficiency, and safety factor to obtain an initial distribution strategy. Then, a dynamic adjustment mechanism is introduced to the initial distribution strategy, allowing the distribution strategy to be updated in real time with the latest weather changes and ocean current directions, ensuring that the effectiveness and accuracy of distribution can be maintained even in a complex and changing environment. Finally, the interactive learning process between the reinforcement learning model and the actual distribution situation is used to automatically adjust the distribution route planning, and the optimal distribution route planning is obtained based on the latest environmental changes and distribution feedback.

[0129] Here is a specific example:

[0130] As another example, consider a marine emergency scenario where a cruise ship sends out an emergency medical distress signal reporting a passenger suffering a heart attack. After receiving the signal, the system quickly generates a detailed disease prediction report and initiates the delivery process of special medical supplies (such as a defibrillator and drugs).

[0131] First, the system uses the comprehensive environmental factor assessment module to obtain and analyze key environmental parameters of weather changes and ocean current direction (such as wind speed, temperature, humidity, air pressure, rainfall, ocean current speed, etc.) in real time, and generates a report on environmental conditions that affect delivery route planning. This report describes in detail the weather conditions and ocean current dynamics in the current and future period, providing an important reference for delivery route planning.

[0132] Next, based on the environmental condition report, the system applied a reinforcement learning algorithm to create a multi-objective optimization problem model that included transportation costs, time efficiency, and safety factors. This model not only takes into account the cost-effectiveness of distribution, but also takes into account time and safety, ensuring that the best distribution strategy is found under multiple constraints. By simulating different distribution scenarios, the system generated a preliminary distribution plan, namely the initial distribution strategy, and determined the best route from the nearest medical supply storage point to the cruise ship.

[0133] In order to cope with the complex marine environment, the system introduces a dynamic adjustment mechanism that allows the distribution strategy to be updated in real time with the latest weather changes and current directions. For example, if the forecast shows that strong winds or large waves are about to occur, the system will automatically adjust the distribution route and choose a safer but slightly slower route to ensure the safe delivery of materials. This dynamic adjustment mechanism ensures that the effectiveness and accuracy of distribution can be maintained even in complex and changing environments, generating continuously optimized distribution route solutions.

[0134] Finally, the system uses the interactive learning process of the reinforcement learning model and the actual delivery situation to automatically adjust the delivery route planning. When drones or rapid response ships are performing delivery tasks, the system continuously optimizes the delivery route based on the latest environmental changes (such as real-time wind speed and current intensity) and delivery feedback (such as actual speed and fuel consumption). Through continuous interactive learning, the system finally obtains the optimal delivery route planning, ensuring that the defibrillator and drugs can be safely delivered to the cruise ship in the shortest time, winning precious treatment time for patients.

[0135] Through these measures, the system not only improves the effectiveness and accuracy of special medical supplies distribution route planning, but also maintains the flexibility and reliability of distribution in complex and changeable marine environments, significantly enhancing the success rate of maritime emergency missions.

[0136] In order to solve the problem that it is difficult to balance transportation cost, time efficiency and safety factor in a complex environment, in some embodiments, based on the environmental condition report, a reinforcement learning algorithm is applied to create a multi-objective optimization problem model including transportation cost, time efficiency and safety factor to obtain an initial distribution strategy, including:

[0137] Using the environmental condition report, the key environmental parameters of weather changes and ocean current direction are deeply analyzed and processed to obtain the analysis results of factors affecting the distribution route selection and material delivery time; based on the factor analysis results, multiple optimization goals are defined, including minimizing transportation costs, maximizing time efficiency, and ensuring the highest safety factor, to generate a multi-objective optimization framework; applying the reinforcement learning algorithm, according to the multi-objective optimization framework, a multi-objective optimization problem model that can handle the trade-off relationship between different optimization goals is constructed, and the goal definition and weight allocation in the multi-objective optimization framework are used as the basis for model construction; a simulation environment is constructed to simulate different distribution scenarios, provide a training platform for the multi-objective optimization problem model, and use the multi-objective optimization model in the simulation environment; through a large number of simulation experiments, the reinforcement learning model is allowed to continuously try distribution route selection in the simulation environment, and the trade-off relationship between each goal is adjusted according to feedback, gradually converge to the best strategy combination, and obtain a preliminary optimized distribution plan; based on the preliminary optimized distribution plan, the final adjustment is made in combination with the latest environmental conditions and actual distribution needs to generate an initial distribution strategy.

[0138] In this embodiment, the factor analysis results use the environmental condition report to deeply analyze the key environmental parameters of weather changes and ocean current directions (such as wind speed, temperature, humidity, air pressure, rainfall, ocean current speed, etc.) to obtain the factor analysis results that affect the distribution path selection and material delivery time. These factor analysis results provide important information about the optimal distribution path under different environmental conditions.

[0139] The multi-objective optimization framework is based on the factor analysis results, defines multiple optimization objectives, including minimizing transportation costs, maximizing time efficiency, and ensuring the highest safety factor, and generates a multi-objective optimization framework. Each optimization objective has a clear definition and weight assignment to guide the subsequent model construction and optimization process. The multi-objective optimization problem model is to apply the reinforcement learning algorithm, and according to the multi-objective optimization framework, construct a multi-objective optimization problem model that can handle the trade-off relationship between different optimization objectives. This model not only considers the optimal solution of a single objective, but also finds the global optimal solution by adjusting the weights between the objectives. The simulation environment is to build a simulation environment, simulate different distribution scenarios, and provide a training platform for the multi-objective optimization problem model. The simulation environment can simulate various possible distribution situations, including different weather conditions, ocean current directions, traffic congestion and other factors, to help the model better adapt to the actual environment. The best strategy combination is through a large number of simulation experiments, in which the reinforcement learning model continuously tries to select the distribution path in the simulation environment, and adjusts the trade-off relationship between the objectives according to feedback, and gradually converges to the best strategy combination. This process enables the model to find the most suitable distribution path in a complex environment. The initial delivery strategy is based on the above-mentioned preliminary optimized delivery plan, combined with the latest environmental conditions and actual delivery needs, and then the final adjustment is made to generate the initial delivery strategy. This strategy not only takes into account the changes in the current environment, but also makes fine adjustments according to the actual delivery needs to ensure its applicability and effectiveness.

[0140] In the embodiment of the present application, first, the environmental condition report is used to deeply analyze the key environmental parameters of weather changes and ocean current direction, and the analysis results of factors affecting the selection of distribution routes and the delivery time of materials are obtained. Secondly, based on the factor analysis results, multiple optimization objectives are defined, including minimizing transportation costs, maximizing time efficiency, and ensuring the highest safety factor, a multi-objective optimization framework is generated, and the definition and weight distribution of each objective are clarified. Then, a reinforcement learning algorithm is applied, and a multi-objective optimization problem model that can handle the trade-off relationship between different optimization objectives is constructed according to the multi-objective optimization framework. Further, a simulation environment is constructed to simulate different distribution scenarios, provide a training platform for the multi-objective optimization problem model, and use the multi-objective optimization model in the simulation environment. Furthermore, through a large number of simulation experiments, the reinforcement learning model is allowed to continuously try the distribution route selection in the simulation environment, and the trade-off relationship between the objectives is adjusted according to the feedback, and gradually converges to the optimal strategy combination to obtain a preliminary optimized distribution plan. Finally, based on the preliminary optimized distribution plan, combined with the latest environmental conditions and actual distribution needs, a final adjustment is made to generate an initial distribution strategy.

[0141] Here is a specific example:

[0142] As another example, consider a marine emergency scenario where a cruise ship sends out an emergency medical distress signal reporting a passenger suffering a heart attack. After receiving the signal, the system quickly generates a detailed disease prediction report and initiates the delivery process of special medical supplies (such as a defibrillator and drugs).

[0143] First, the system uses the comprehensive environmental factor assessment module to obtain and analyze key environmental parameters of weather changes and ocean current direction (such as wind speed, temperature, humidity, air pressure, rainfall, ocean current speed, etc.) in real time, and generates a report on environmental conditions that affect delivery route planning. This report describes in detail the weather conditions and ocean current dynamics in the current and future period, providing an important reference for delivery route planning.

[0144] Next, the system conducted an in-depth analysis of these key environmental parameters and obtained the analysis results of factors that affect the selection of distribution routes and the delivery time of materials. For example, strong winds may cause unstable flight of drones, large waves may increase the risk of ship navigation, and low temperatures may affect equipment performance. These factor analysis results provide a basis for subsequent optimization.

[0145] Based on the analysis results of these factors, the system defines multiple optimization goals, including minimizing transportation costs, maximizing time efficiency, and ensuring the highest safety factor, generating a multi-objective optimization framework. Each optimization goal is given a corresponding weight to reflect its importance. For example, the safety factor may be given a higher weight because the patient's life safety is the most important.

[0146] Then, the system applies the reinforcement learning algorithm and, based on the multi-objective optimization framework, builds a multi-objective optimization problem model that can handle the trade-offs between different optimization objectives. This model not only considers the optimal solution for a single objective, but also finds the global optimal solution by adjusting the weights between the objectives. To train this model, the system builds a simulation environment that simulates different delivery scenarios, including sunny days, heavy rain, strong winds, and big waves. The simulation environment provides the model with rich training data, enabling it to better adapt to the actual environment.

[0147] In the simulation environment, the reinforcement learning model continuously tries to select delivery routes and adjusts the trade-offs between various objectives based on feedback. For example, if a route is the fastest but too risky, the model will automatically adjust to select a safer but slightly slower route. After a large number of simulation experiments, the model gradually converges to the best strategy combination and obtains a preliminary optimized delivery solution.

[0148] Finally, the system made final adjustments based on the initially optimized distribution plan, combined with the latest environmental conditions (such as real-time wind speed and current intensity) and actual distribution needs (such as the type and quantity of materials), and generated an initial distribution strategy. For example, if the latest forecast shows that strong winds are about to occur, the system will re-evaluate and choose a safer path to ensure that the materials can reach the cruise ship safely. Through continuous interactive learning, the system finally determined the optimal distribution path to ensure that the defibrillator and drugs can be safely delivered to the cruise ship in the shortest time, winning precious treatment time for the patients.

[0149] Through these measures, the system not only improves the adaptability and accuracy of the multi-objective optimization problem model in complex environments, but also maximizes the distribution efficiency while ensuring safety, significantly enhancing the success rate of maritime emergency rescue missions.

[0150] In order to solve the problems of real-time and accuracy of data collection and analysis at the emergency scene, in some embodiments, according to the optimized communication link, vital sign monitoring data and specific key environmental indicators are collected and transmitted back to the medical command center in real time, and deep learning algorithms are applied to perform pattern recognition and trend prediction, generate potential risk warnings, and form updated medical needs, including:

[0151] By using the optimized communication link, the vital sign monitoring data and specific key environmental indicators at the emergency scene are collected and processed to obtain the encoded monitoring data, wherein the specific key environmental indicators include ambient temperature, humidity, air pressure, wind speed, rainfall and ocean current speed; based on the encoded monitoring data, the encoded monitoring data is transmitted and processed in real time through the optimized communication link to obtain a data stream transmitted to the medical command center in real time; according to the data stream transmitted to the medical command center in real time, the deep learning algorithm deployed in the medical command center is applied to analyze and process the vital sign monitoring data and specific key environmental indicators to obtain a pattern recognition result, wherein the specific key environmental indicators include ambient temperature, humidity, air pressure, wind speed, rainfall and ocean current speed; based on the pattern recognition result, the deep learning model is used to predict the development trend of the patient's condition and generate a trend prediction report; according to the trend prediction report, the system automatically generates a potential risk warning for each patient, and obtains warning information including current risk factors and future risk prompts; based on the warning information, the medical command center re-evaluates and adjusts the emergency plan to form updated medical needs.

[0152] In this embodiment, the encoded monitoring data is obtained by collecting and processing the vital signs monitoring data (such as heart rate, blood pressure, blood oxygen saturation, etc.) and specific key environmental indicators (such as ambient temperature, humidity, air pressure, wind speed, rainfall and ocean current speed) at the emergency scene using the optimized communication link. After encoding, these data are convenient for efficient transmission and subsequent processing. The data stream transmitted to the medical command center in real time is based on the encoded monitoring data. Through the optimized communication link, the encoded monitoring data is transmitted and processed in real time to obtain the data stream transmitted to the medical command center in real time. This process ensures the immediacy and reliability of the data, so that the medical command center can quickly obtain the latest information. The pattern recognition result is based on the data stream transmitted to the medical command center in real time, and the deep learning algorithm deployed in the medical command center is applied to analyze and process the vital signs monitoring data and specific key environmental indicators to obtain the pattern recognition result. These results not only reflect the current health status, but also reveal possible abnormal patterns. The trend prediction report is based on the pattern recognition result, and the development trend of the patient's condition is predicted and processed using a deep learning model to generate a trend prediction report. The report describes in detail the possible future changes in the patient's condition and provides a scientific basis for medical decision-making. Based on the trend forecast report, the system automatically generates a potential risk warning for each patient, and obtains warning information including current risk factors and future risk prompts. These warning information helps medical experts predict and respond to potential risks in advance. The updated medical needs are based on the warning information. The medical command center re-evaluates and adjusts the emergency plan to form updated medical needs. The new medical needs take into account the latest changes in the condition and environmental conditions, ensuring the effectiveness and pertinence of the treatment measures.

[0153] First, the optimized communication link is used to collect and process the vital signs monitoring data and specific key environmental indicators at the emergency scene to obtain the encoded monitoring data. Secondly, based on the encoded monitoring data, the encoded monitoring data is transmitted and processed in real time through the optimized communication link to obtain a data stream transmitted to the medical command center in real time. Then, according to the data stream transmitted to the medical command center in real time, the deep learning algorithm deployed in the medical command center is applied to analyze and process the vital signs monitoring data and specific key environmental indicators to obtain a pattern recognition result. Further, based on the pattern recognition result, the deep learning model is used to predict the development trend of the patient's condition and generate a trend prediction report. Furthermore, according to the trend prediction report, the system automatically generates a potential risk warning for each patient, and obtains warning information including current risk factors and future risk prompts. Finally, based on the warning information, the medical command center re-evaluates and adjusts the emergency plan to form updated medical needs.

[0154] Here is a specific example:

[0155] As another example, consider a marine emergency scenario where a cruise ship sends an emergency medical distress signal reporting a passenger suffering a heart attack. After receiving the signal, the system quickly establishes an optimized satellite communication link and begins to collect real-time vital sign monitoring data and certain key environmental indicators at the emergency scene.

[0156] First, the system uses the optimized communication link to collect and process the vital signs monitoring data (such as heart rate, blood pressure, blood oxygen saturation, etc.) and specific key environmental indicators (such as ambient temperature, humidity, air pressure, wind speed, rainfall and ocean current speed) at the emergency scene to obtain coded monitoring data. After being coded, these data are easy to transmit efficiently and process later.

[0157] Next, the system transmits the encoded monitoring data in real time through the optimized communication link, ensuring that the medical command center can quickly obtain the latest information. The real-time data stream includes changes in the patient's vital signs and environmental conditions, providing rich data support for subsequent analysis.

[0158] At the medical command center, the deployed deep learning algorithm performs detailed analysis and processing on the real-time transmitted data streams, and obtains pattern recognition results. For example, the algorithm detects abnormal fluctuations in the patient's heart rate, and also finds that the low ambient temperature may cause the patient's body temperature to drop. These pattern recognition results not only reflect the current health status, but also reveal possible abnormal patterns.

[0159] Based on these pattern recognition results, the system used a deep learning model to predict the development trend of the patient's condition and generated a trend prediction report. The report pointed out that if the current low temperature environment continues, the patient may be at risk of hypothermia and recommended immediate warming measures. In addition, the model also predicted the possibility that the patient's heart function may deteriorate in the next few hours, reminding medical experts to be prepared.

[0160] Based on the trend forecast report, the system automatically generates a potential risk warning for the patient, including current risk factors (such as abnormal heart rate fluctuations, low temperature environment) and possible future risk warnings (such as hypothermia, deterioration of heart function). These warning information helps medical experts predict and respond to potential risks in advance.

[0161] Finally, based on the warning information, the medical command center re-evaluated and adjusted the emergency plan and formed updated medical needs. For example, the demand for warming equipment was increased, and additional medical supplies (such as heating blankets and thermal clothing) were arranged to ensure that patients can receive appropriate treatment in the shortest time. At the same time, the system also adjusted the drone delivery route, choosing a safer but slightly slower route to ensure the safe delivery of supplies.

[0162] Through these measures, the system not only improves the real-time and accuracy of data collection and analysis at the emergency scene, but also maintains the effectiveness and pertinence of rescue measures in complex and changing environments, significantly enhancing the success rate of maritime emergency missions.

[0163] In order to solve the effectiveness and adaptability of the emergency site material distribution strategy in the face of potential risks and updated needs, in some embodiments, based on the potential risk warning and updated medical needs, the optimal distribution path planning is re-evaluated and optimized to generate an updated optimal distribution path planning to ensure that the materials can be delivered to the emergency site in a timely and accurate manner, including:

[0164] Utilize the potential risk warning to analyze and process the risk factors at the emergency site to obtain a detailed risk assessment report; based on the risk assessment report and the updated medical needs, re-evaluate the current medical supplies needs and generate an up-to-date list of medical supplies needs; based on the up-to-date list of medical supplies needs, check whether the existing optimal distribution route planning needs to be adjusted to adapt to new demand changes, and preliminarily determine the aspects that need to be optimized; based on the preliminarily determined optimization aspects, consider the latest environmental conditions and re-plan the optimal distribution route to ensure that the material distribution strategy meets the latest medical needs and risk warnings; through continuous updating and adjustment, ultimately generate an updated optimal distribution route planning to ensure that special medical supplies can be delivered to the emergency site in a timely and accurate manner.

[0165] In this embodiment, the detailed risk assessment report uses the potential risk warning to analyze and process the risk factors at the emergency site to obtain a detailed risk assessment report. The report not only contains the current risk factors (such as ambient temperature, humidity, air pressure, wind speed, rainfall and current speed), but also predicts the risks that may occur in the future and provides corresponding response suggestions. The latest medical material demand list is based on the risk assessment report and the updated medical needs, re-evaluates the current medical material needs, and generates the latest medical material demand list. This list clarifies the types, quantities and urgency of the required materials, and provides clear guidance for subsequent distribution planning. The preliminary optimization aspects are based on the latest medical material demand list, check whether the existing optimal distribution path planning needs to be adjusted to adapt to the new demand changes, and preliminarily determine the aspects that need to be optimized. For example, it may be necessary to increase or decrease the transportation volume of certain materials, or choose a safer but slightly slower distribution path. Replanning the optimal distribution path is based on the preliminary optimization aspects, considering the latest environmental conditions (such as weather forecasts, current direction, etc.), and replanning the optimal distribution path to ensure that the material distribution strategy meets the latest medical needs and risk warnings. This process ensures the safety and timeliness of the distribution route. The updated optimal distribution route planning is generated through continuous updating and adjustment, ensuring that special medical supplies can be delivered to the emergency site in a timely and accurate manner. This plan not only takes into account the latest changes in demand, but also combines real-time environmental data to improve the reliability and flexibility of distribution.

[0166] In an embodiment of the present application, first, the potential risk warning is used to analyze and process the risk factors at the emergency site to obtain a detailed risk assessment report. Secondly, based on the risk assessment report and the updated medical needs, the current medical supplies needs are re-evaluated to generate an up-to-date list of medical supplies needs. Then, based on the up-to-date list of medical supplies needs, check whether the existing optimal distribution route planning needs to be adjusted to adapt to new changes in demand, and preliminarily determine the aspects that need to be optimized. Furthermore, based on the preliminarily determined optimization aspects, the optimal distribution route is re-planned taking into account the latest environmental conditions to ensure that the material distribution strategy meets the latest medical needs and risk warning conditions. Finally, through continuous updating and adjustment, an updated optimal distribution route plan is ultimately generated to ensure that special medical supplies can be delivered to the emergency site in a timely and accurate manner.

[0167] Here is a specific example:

[0168] As another example, consider a maritime first aid scenario where a cruise ship sends out an emergency medical distress signal reporting a passenger suffering a heart attack. After receiving the signal, the system quickly generates a detailed disease prediction report and initiates the delivery process for special medical supplies (such as defibrillators and drugs). As the disease progresses and environmental conditions change, the system needs to continuously optimize the delivery path to ensure that the supplies are delivered in a timely and accurate manner.

[0169] First, the system used potential risk warnings to analyze and process the risk factors at the emergency scene and obtained a detailed risk assessment report. The report pointed out the risk of hypothermia due to low temperatures at night, and that future strong winds and high waves may affect the safety of drone and ship delivery. This information helps the system fully understand the complex environment of the emergency scene.

[0170] Next, the system reassessed the current medical supply needs based on the risk assessment report and the updated medical needs, and generated an updated list of medical supply needs. The list added the need for warming equipment (such as heating blankets and thermal clothing) and adjusted the quantity and type of medicines to better respond to the progression of the patient's condition and environmental changes.

[0171] Based on the latest list of medical supplies requirements, the system checked the existing optimal distribution route planning and preliminarily identified areas that needed to be optimized. For example, considering the strong winds and waves in the coming days, the system decided to choose a safer but slightly slower distribution route to ensure the safe delivery of supplies.

[0172] Then, based on the above-preliminary optimization aspects, the system replanned the optimal delivery route taking into account the latest environmental conditions (such as weather forecasts, ocean current direction, etc.). The system simulated the delivery time, cost and safety under different routes and finally selected the best route. For example, a route that bypasses the storm area may be a little longer, but it can avoid the risks brought by bad weather.

[0173] Finally, through continuous updating and adjustment, the system finally generated an updated optimal delivery route plan to ensure that special medical supplies can be delivered to the emergency site in a timely and accurate manner. The system continuously monitors environmental changes and delivery progress, dynamically adjusts the route, and ensures that the delivery task is completed smoothly. For example, when it finds that the weather in a certain area has improved, the system will automatically switch back to a faster route to shorten the delivery time.

[0174] Through these measures, the system not only improves the effectiveness and adaptability of the material distribution strategy at the emergency site, but also maintains the safety and reliability of distribution in a complex and changing environment, significantly enhancing the success rate of maritime emergency missions.

[0175] This application takes into account that in the marine emergency scenario, the distribution route planning of special medical supplies faces many challenges, such as complex weather changes, ocean current direction and other environmental factors. The traditional single-objective optimization method is difficult to take into account the three key indicators of transportation cost, time efficiency and safety factor at the same time. Therefore, a multi-objective optimization method is needed to balance these conflicting goals and ensure the optimal distribution strategy in a complex and changing environment. Therefore, a new optional solution is proposed, which includes:

[0176] Based on the environmental condition report, a reinforcement learning algorithm is applied to create a multi-objective optimization problem model including transportation cost, time efficiency, and safety factor to obtain an initial distribution strategy, including:

[0177] Using the environmental condition report, the key environmental parameters of weather changes and ocean current direction are deeply analyzed and processed to obtain the analysis results of factors affecting the distribution route selection and material delivery time. The key environmental parameters include wind speed, temperature, rainfall and ocean current speed;

[0178] Based on the factor analysis results, multiple optimization objectives are defined to generate a multi-objective optimization framework. The generated optimization objective definitions and weight distribution will serve as data input for the next step. The optimization objectives include minimizing the transportation cost C t , maximize time efficiency E t And ensure the highest safety factor F t , the weight of each target is w C 、w E and w F , and satisfies ∑w i =1;

[0179] The weight w is expressed by the following formula C 、w E and w F :

[0180]

[0181] Among them, α, β and γ are adjustment factors used to adjust the importance of different goals; C t is to minimize transportation costs; E t is to maximize time efficiency; F t is to ensure the highest safety factor; e is the natural base;

[0182] Applying a reinforcement learning algorithm, based on the multi-objective optimization framework, constructing a multi-objective optimization problem model capable of handling the trade-off relationship between different optimization objectives; the multi-objective optimization problem model uses the objectives and their weights in the multi-objective optimization framework as a basis for construction;

[0183] The multi-objective optimization problem is expressed by the following formula:

[0184]

[0185] Among them, C d is the transportation cost, E d is the time efficiency, F d is the safety factor; δ, ∈ and ζ are the exponential factors of transportation cost, time efficiency and safety factor, respectively, which are used to adjust the impact of each target;

[0186] Constructing a simulation environment to simulate different delivery scenarios, providing a training platform for the multi-objective optimization problem model, using information obtained from the multi-objective optimization problem model to guide the design of the simulation environment, and ensuring that the multi-objective optimization problem model can be trained under conditions close to reality;

[0187] Through a large number of simulation experiments, the reinforcement learning model is allowed to continuously try to select the delivery path in the simulation environment, and the trade-off relationship between the various objectives is adjusted according to the feedback, gradually converging to the optimal strategy combination, and obtaining a preliminary optimized delivery plan;

[0188] The updated policy π is calculated for each iteration t using the following formula: t+1 :

[0189]

[0190] Among them, C d (π), E d (π), F d (π) are the transportation cost, time efficiency and safety factor when adopting strategy π.

[0191] The following is a detailed explanation of each parameter:

[0192] Key environmental parameters:

[0193] Wind speed: affects the speed and stability of drones or ships. It is obtained in real time through weather stations or satellite data.

[0194] Temperature: May affect device performance (such as battery life). Obtained through field sensors or weather forecast data.

[0195] Precipitation: May reduce visibility or increase risk of skidding. Obtained from weather radar or weather forecast data.

[0196] Ocean current speed: affects the course and speed of ships at sea. Obtained through ocean monitoring systems or historical data analysis.

[0197] C t:It is to minimize transportation costs, reduce fuel consumption, labor costs, etc. Obtained through historical data statistics of the logistics management system.

[0198] E t :Maximize time efficiency and shorten delivery time as much as possible. Calculated based on historical delivery time and traffic conditions data.

[0199] F t :To ensure the highest safety factor and ensure that there are no accidents during the delivery process. Determined by risk assessment model and historical accident data.

[0200] w C : The weight of the transportation cost, indicating the importance of minimizing the transportation cost. It is set based on historical data and specific application scenarios.

[0201] w E : The weight of time efficiency, indicating the importance of maximizing time efficiency. Set according to the urgency of the task.

[0202] w F : The weight of the safety factor, indicating the importance of ensuring the highest safety factor. Set according to the mission risk assessment.

[0203] C d : It is the transportation cost, which indicates the actual transportation cost when a certain distribution strategy is adopted, including fuel consumption, labor cost, etc. It is obtained through historical data statistics of the logistics management system.

[0204] E d : It is the time efficiency, which indicates the delivery time when a certain delivery strategy is adopted, that is, the total time required from the starting point to the end point. It is calculated by the GPS positioning system or the path planning algorithm in the simulation environment.

[0205] F d : is the safety factor, which indicates the safety of a distribution strategy. It is usually a value between 0 and 1, with 1 indicating complete safety. It is determined through risk assessment models and historical accident data.

[0206] α: Adjusts the importance of transportation costs. Set according to specific application scenarios. For example, when resources are limited, the α value can be increased to pay more attention to cost control.

[0207] β: Adjusts the importance of time efficiency. In emergency situations, the value can be increased to place more emphasis on fast response.

[0208] γ: adjusts the importance of the safety factor. For high-risk tasks, the γ value should be increased to ensure safety priority.

[0209] δ: Adjusts the impact of shipping costs. Usually set to a positive value to emphasize cost savings.

[0210] ∈: Adjust the impact of time efficiency. Usually set to a positive value to emphasize fast response.

[0211] ζ: Adjusts the impact of the safety factor. Usually set to a negative value to emphasize safety.

[0212] π: represents a delivery route selection scheme, including the starting point, passing points, end point, speed of each section, route and other information. It is obtained through iterative optimization of the reinforcement learning algorithm. After each iteration, it is updated to the optimal strategy π t+1 .

[0213] C d (π): is the transportation cost, which indicates the actual transportation cost when a certain distribution strategy π is adopted, including fuel consumption, labor cost, etc. It is obtained through historical data statistics of the logistics management system and adjusted in combination with current environmental conditions (such as wind speed and ocean current speed).

[0214] E d (π): is the time efficiency, which represents the delivery time when a certain delivery strategy T is adopted, that is, the total time required from the starting point to the end point. It is calculated by the GPS positioning system or the path planning algorithm in the simulation environment, and takes into account the real-time traffic conditions and weather effects.

[0215] F d (π): is the safety factor, which indicates the safety of a distribution strategy π. It is usually a value between 0 and 1, with 1 indicating complete safety. It is determined through risk assessment models and historical accident data, and dynamically adjusted in combination with current environmental conditions (such as wind speed and rainfall).

[0216] The following is an introduction to the design reasons of each sub-item:

[0217] This sub-item represents the contribution of transportation costs in the multi-objective optimization problem. By introducing the exponential factor δ, the impact of transportation costs on the overall goal can be flexibly adjusted. When δ is large, the impact of transportation costs is amplified; otherwise, it is reduced. This approach allows the system to flexibly adjust the importance of cost control in different scenarios.

[0218] This sub-item represents the contribution of time efficiency in multi-objective optimization problems. Use the inverse of time efficiency This is because the shorter the time, the larger its reciprocal, thus highlighting the importance of time efficiency in the optimization process. By introducing the exponential factor ∈, the impact of time efficiency on the overall goal can be flexibly adjusted. When ∈ is large, the impact of time efficiency is amplified; otherwise, it is reduced.

[0219] This sub-item represents the contribution of the safety factor in the multi-objective optimization problem. Because the safety factor F d The closer it is to 1 (i.e., safer), the smaller its value is, thus giving more priority to safety during the optimization process. By introducing an exponential factor, the impact of the safety factor on the overall goal can be flexibly adjusted. When ζ is larger, the impact of safety is amplified; otherwise, it is reduced.

[0220] The addition of the three sub-items allows for a linear combination of different objectives, making them easier to understand and achieve. The weight and exponential factor of each sub-item ensure that the importance of each objective is reflected in the final optimization result. Transport cost, time efficiency, and safety factor are usually conflicting objectives. By adding them together, the importance of each objective can be flexibly adjusted in different application scenarios to find a compromise solution.

[0221] The purpose of the entire formula is to build a multi-objective optimization framework that can balance the three conflicting objectives of transportation cost, time efficiency, and safety factor in a complex and changing environment. By introducing weights and exponential factors, the model can flexibly adjust the importance of each objective according to the specific application scenario. At the same time, using the reinforcement learning algorithm, different delivery path options are continuously tried in the simulation environment, and the trade-off relationship between the objectives is adjusted based on feedback, gradually converging to the optimal strategy combination.

[0222] w C ·C d (π) δ : This sub-item represents the contribution of transportation cost in the multi-objective optimization problem. By introducing the exponential factor δ, the impact of transportation cost on the overall goal can be flexibly adjusted. When δ is large, the impact of transportation cost is amplified; otherwise, it is reduced. This approach allows the system to flexibly adjust the importance of cost control in different scenarios.

[0223] This sub-item represents the contribution of time efficiency in multi-objective optimization problems. Use the inverse of time efficiency This is because the shorter the time, the larger its reciprocal, thus highlighting the importance of time efficiency in the optimization process. By introducing the exponential factor ∈, the impact of time efficiency on the overall goal can be flexibly adjusted. When ∈ is large, the impact of time efficiency is amplified; otherwise, it is reduced.

[0224] w F ·(1-F d (π)ζ): This sub-item represents the contribution of the safety factor in the multi-objective optimization problem. Because the safety factor F dThe closer it is to 1 (i.e., safer), the smaller its value is, thus giving more prominence to the priority of safety during the optimization process. By introducing the exponential factor ζ, the impact of the safety factor on the overall goal can be flexibly adjusted. When ζ is larger, the impact of safety is amplified; otherwise, it is reduced.

[0225] The addition of the three sub-items allows for a linear combination of different objectives, making them easier to understand and achieve. The weight and exponential factor of each sub-item ensure that the importance of each objective is reflected in the final optimization result. Transport cost, time efficiency, and safety factor are usually conflicting objectives. By adding them together, the importance of each objective can be flexibly adjusted in different application scenarios to find a compromise solution.

[0226] The purpose of the entire formula is to build a multi-objective optimization framework that can balance the three conflicting objectives of transportation cost, time efficiency, and safety factor in a complex and changing environment. By introducing weights and exponential factors, the model can flexibly adjust the importance of each objective according to the specific application scenario. At the same time, using the reinforcement learning algorithm, different delivery path options are continuously tried in the simulation environment, and the trade-off relationship between the objectives is adjusted based on feedback, gradually converging to the optimal strategy combination.

[0227] Here is a specific example:

[0228] Assume that in a specific marine emergency scenario, a cruise ship sends out an emergency medical distress signal, reporting that a passenger has suffered a heart attack. After receiving the signal, the system quickly generates a detailed medical prediction report and initiates the distribution process of special medical supplies (such as defibrillators and drugs). In order to ensure that the supplies can be safely delivered to the cruise ship in the shortest time, the system applies the above multi-objective optimization problem model for path planning.

[0229] The following are the parameter settings:

[0230] Adjustment factors: α = 0.1, β = 0.2, γ = 0.3; Exponential factors: δ = 1, ∈ = 1, ( = 1;

[0231] Initial weight: W C =0.4,W E =0.3,W F =0.3;

[0232] Assume the current environmental conditions are as follows:

[0233] Wind speed: 15m / s; Temperature: 15℃; Rainfall: 5mm / h; Current speed: 2knots;

[0234] Calculate weights:

[0235] Assume that the transportation cost is C t=500 yuan, time efficiency E t =6 hours, safety factor F t =0.9 (full score is 1):

[0236]

[0237] After calculation, the updated weight value is obtained:

[0238] W C ≈0.0001;

[0239] W E ≈0.748;

[0240] W F ≈0.252;

[0241] Construct a multi-objective optimization problem model:

[0242] Assume that in a certain iteration, the transportation cost C under strategy π is d (π) = 450 yuan, time efficiency E d (π) = 5 hours, safety factor F d (π)=0.95

[0243] The optimization objective function is min(0.0001·450 1 +0.748 (1 / 5) 1 +0.252·(1-0.95 1 ))

[0244] The calculation results are:

[0245] min(0.045+0.1496+0.126)≈0.321

[0246] Update strategy:

[0247] Update the strategy π by the following formula t+1 :

[0248]

[0249] Suppose that after many iterations, the system finds a new route that reduces the transportation cost to 400 yuan, increases the time efficiency to 4 hours, and keeps the safety factor unchanged at 0.95.

[0250] Then the new optimization objective function is min(0.0001·400 1 +0.748·(1 / 4) 1 +0.252·(1-0.95 1 ))

[0251] The calculation results are:

[0252] min(0.04+0.187+0.126)≈0.353

[0253] Since the optimization target value of the new route is slightly higher than the previous result, it means that although the transportation cost and time efficiency have improved, they are not significantly better than the previous route overall. Therefore, the system may choose to keep the original route or further fine-tune it to find a better solution.

[0254] Through the above calculations, the following conclusions can be drawn: the initial weights are set to 0.4, 0.3 and 0.3 for transportation cost, time efficiency and safety factor respectively. By calculating the updated weights, it is found that the importance of time efficiency is significantly amplified, while the impact of transportation cost becomes very small. This shows that in the current environment, time efficiency is the most important optimization goal, followed by safety factor, and finally transportation cost. Although the new path reduces transportation costs and improves time efficiency, the overall optimization target value is not significantly reduced due to the high weight of time efficiency. This suggests that in practical applications, it may be necessary to re-evaluate the weights of each goal or explore other potential paths to find the true optimal solution. Through continuous iteration and feedback adjustment, the system can gradually approach the optimal path. In this process, the simulation environment provides an important training platform to help the system better adapt to the complex and changeable marine environment.

[0255] In summary, this multi-objective optimization problem model not only effectively solves the problem of special medical supplies distribution path planning in maritime emergency scenarios, but also maintains the safety and reliability of distribution in complex and changing environments, significantly enhancing the success rate of maritime emergency missions.

[0256] This application takes into account that in emergency medical rescue scenarios, especially in maritime emergency rescue missions, real-time collection and analysis of vital sign monitoring data and specific key environmental indicators are essential for effective treatment of patients. Traditional data analysis methods often have difficulty coping with large, complex and dynamically changing data streams, and cannot provide accurate disease prediction and risk warning. Therefore, an efficient and intelligent data processing method is needed to support remote medical decision-making, so a new optional solution is proposed, which includes:

[0257] According to the optimized communication link, vital sign monitoring data and specific key environmental indicators are collected and transmitted back to the medical command center in real time, and deep learning algorithms are applied to perform pattern recognition and trend prediction to generate potential risk warnings and form updated medical needs. The specific key environmental indicators include ambient temperature, humidity, air pressure, wind speed, rainfall and ocean current speed, including:

[0258] By using the optimized communication link, vital sign monitoring data and specific key environmental indicators at the emergency scene are collected and processed to obtain encoded monitoring data, wherein the specific key environmental indicators include ambient temperature, humidity, air pressure, wind speed, rainfall and ocean current speed;

[0259] Based on the encoded monitoring data, the encoded monitoring data is transmitted and processed in real time through the optimized communication link to obtain a data stream transmitted to the medical command center in real time, ensuring the security and reliability of data transmission, and using encryption technology and redundancy mechanism to prevent data loss or leakage;

[0260] According to the data stream D transmitted to the medical command center in real time, a deep learning algorithm deployed in the medical command center is applied to analyze and process the vital sign monitoring data and specific key environmental indicators to obtain a pattern recognition result R;

[0261] The pattern recognition result R is calculated by the following formula:

[0262]

[0263] Where f(·) is the feature extraction function in the deep learning model; θ represents the model parameters; w i Each data point D i The weight of 1 is an exponential factor that adjusts the impact of each data point; R is the result obtained by the deep learning model after analyzing and processing the vital signs monitoring data and specific key environmental indicators;

[0264] Based on the pattern recognition result R, a deep learning model is used to predict the development trend of the patient's condition and generate a trend prediction report T pred ;

[0265] Calculate the trend forecast report T using the following formula pred :

[0266]

[0267] Where g(·) is the prediction function in the deep learning model; φ represents the model parameters; H t Represents historical data; β 1 is the time decay factor; used to adjust the influence weights at different time points; R t represents the pattern recognition result at time point t; f pred It is a trend forecast report;

[0268] According to the trend forecast report Tp red, the system automatically generates a potential risk warning W for each patient, and obtains warning information including current risk factors and possible risk warnings in the future;

[0269] The potential risk warning W is calculated by the following formula:

[0270]

[0271] Where h(·) is the risk assessment function; ψ represents the model parameter; C curr Indicates the current condition; γ 1 and δ 1 are the influencing factors of trend prediction and current conditions respectively; W is the potential risk warning;

[0272] Based on the warning information W, the medical command center re-evaluates and adjusts the emergency plan to form an updated medical demand M new ;

[0273] The updated medical demand M is calculated by the following formula: new :

[0274]

[0275] Where k(·) is the demand assessment function; ω represents the model parameter; D curr Indicates the latest data stream; η j and k are the weights of warning information and the latest data stream respectively; ∈ 1 and λ 1 are the index factors for adjusting their respective impacts; M new is the updated medical need; W j represents the jth risk factor in the warning information; max(0,·) ensures that the calculation result is non-negative, even if the result of the weighted sum is negative, the final M new The value will also be set to 0.

[0276] Here is a specific example:

[0277] Suppose in a specific marine emergency scenario, a cruise ship sends out an emergency medical distress signal, reporting that a passenger has suffered a sudden heart attack. After receiving the signal, the system quickly establishes an optimized satellite communication link and begins to collect real-time vital sign monitoring data (such as heart rate, blood pressure, blood oxygen saturation, etc.) and specific key environmental indicators (such as ambient temperature, humidity, air pressure, wind speed, rainfall, and current speed) at the emergency scene. In order to ensure that the supplies can be safely delivered to the cruise ship in the shortest time, the system applies the above multi-objective optimization problem model for path planning, and also uses deep learning algorithms to predict the development trend of the patient's condition, generate potential risk warnings, and form updated medical needs.

[0278] The following are the parameter settings:

[0279] Exponential factor: α 1 =0.5,∈ 1 =0.6,λ 1 =0.7;

[0280] Weight: w i =[0.2, 0.3, 0.1, 0.1, 0.1, 0.2] corresponds to each data point D i ;

[0281] Time decay factor: β 1 =0.3;

[0282] Impact Factor: γ 1 =0.4,δ 1 =0.2;

[0283] Warning information weight: η j = [0.6, 0.4];

[0284] Latest data flow weight: ζ k = [0.5, 0.5];

[0285] Assume the current environmental conditions are as follows:

[0286] Vital signs monitoring data: heart rate = 90 beats / minute, blood pressure = 120 / 80 mmHg, blood oxygen saturation = 95%

[0287] Specific key environmental indicators: ambient temperature = 15°C, humidity = 80%, air pressure = 1013hPa, wind speed = 15m / s, rainfall = 5mm / h, current speed = 2knots

[0288] The optimized communication link is used to collect and process the vital signs monitoring data and specific key environmental indicators at the emergency scene to obtain coded monitoring data. After encryption and redundancy processing, these data are transmitted to the medical command center in real time through the optimized communication link to ensure the security and reliability of data transmission.

[0289] Calculate the pattern recognition result R:

[0290] Assume that the vital signs monitoring data and specific key environmental indicators received at a certain moment are D i =[90, 120, 95, 15, 80, 1013, 15, 5, 2]

[0291] Then R=0.2·90 0.5 +0.3 120 0.5 +0.1 95 0.5 +0.1 15 0.5 +0.1 80 0.5 +0.2 1013 0.5

[0292] After calculation, the pattern recognition result is R≈22.4

[0293] Assume that the pattern recognition results at the past few time points are R t =[20, 22, 24, 22.4]

[0294] but

[0295] After calculation, we get the trend forecast report T pred ≈23.2

[0296] Assuming the current condition C curr =0.8 (full score is 1);

[0297] but

[0298] After calculation, we get the potential risk warning W≈0.97;

[0299] Forming an updated medical demand formula M new :

[0300] Assume that the early warning information contains two risk factors W j =[0.97, 0.8], the latest data stream D curr，k =[90, 120]

[0301] Then M new =max(0,0.6·0.97 0.6 +0.4 0.8 0.6 +0.5 900.7 +0.5 120 0.7 )≈117.3

[0302] After calculation, the updated medical demand M is obtained. new ≈117.3;

[0303] Through the above calculations, we can draw the following conclusions: The initial pattern recognition result R≈22.4 reflects the comprehensive status of the current patient's vital signs and environmental conditions. This result provides a basis for subsequent trend prediction. Trend prediction report T pred ≈23.2 indicates that the patient's condition may continue to deteriorate in the next few hours. This suggests that medical experts need to prepare higher-level treatment measures in advance. Potential risk warning W≈0.97 indicates that there are currently high risk factors, especially considering the current conditions, the patient's condition is likely to deteriorate. This warning information helps medical experts take preventive measures in time. Updated medical needs M new ≈117.3 indicates that more medical resources and support are needed to cope with possible risks, such as increasing the demand for warming equipment, adjusting the quantity and type of medicines, and ensuring that patients can receive appropriate treatment in the shortest possible time.

[0304] In summary, this system not only improves the real-time and accuracy of data collection and analysis at the emergency scene, but also maintains the effectiveness and pertinence of treatment measures in a complex and changing environment, significantly enhancing the success rate of maritime emergency missions. Through continuous monitoring and dynamic adjustment, the system can continuously optimize the distribution strategy to ensure that special medical supplies can be delivered to the emergency scene in a timely and accurate manner.

[0305] Figure 2 A schematic diagram of the structure of a maritime emergency information transmission system based on satellite communication is provided for an embodiment of the present application. Figure 2 As shown, the system includes:

[0306] An optimization module 21 is constructed to construct a temporary dedicated satellite communication link according to the emergency location information and the preliminary medical condition description in the received emergency medical distress signal, and optimize the parameters of the temporary dedicated satellite communication link to obtain an optimized communication link;

[0307] The reconstruction and analysis module 22 is used to utilize the optimized communication link, combined with augmented reality technology and computer vision algorithms, to perform three-dimensional reconstruction of the on-site environment, to achieve high-definition audio and video calls and real-time data sharing between medical experts and on-site rescue personnel, and to apply a machine learning model to analyze the description of the severity of the disease, predict the progression of the disease, and generate a disease prediction result;

[0308] The optimization processing module 23 is used to optimize the distribution strategy of special medical supplies based on the disease prediction results by using a reinforcement learning algorithm and taking into account factors such as weather changes and ocean current direction, so as to obtain an optimal distribution route plan;

[0309] The collection and formation module 24 is used to collect vital sign monitoring data and specific key environmental indicators according to the optimized communication link, transmit them back to the medical command center in real time, and apply deep learning algorithms to perform pattern recognition and trend prediction, generate potential risk warnings, and form updated medical needs. The specific key environmental indicators include ambient temperature, humidity, air pressure, wind speed, rainfall, and ocean current speed;

[0310] The evaluation and optimization module 25 is used to re-evaluate and optimize the optimal distribution path plan based on the potential risk warning and the updated medical needs, generate an updated optimal distribution path plan, and ensure that the supplies can be delivered to the emergency site in a timely and accurate manner.

[0311] Figure 2 The satellite communication-based marine emergency information transmission system can be implemented Figure 1 The implementation principle and technical effect of the method for transmitting emergency information at sea based on satellite communication described in the illustrated embodiment will not be described in detail. The specific manner in which each module and unit performs operations in the system for transmitting emergency information at sea based on satellite communication in the above embodiment has been described in detail in the embodiment of the method, and will not be described in detail here.

[0312] In one possible design, Figure 2 The satellite communication-based marine emergency information transmission system of the illustrated embodiment can be implemented as a computing device, such as Figure 3 As shown, the computing device may include a storage component 31 and a processing component 32;

[0313] The storage component 31 stores one or more computer instructions, wherein the one or more computer instructions are called and executed by the processing component 32 .

[0314] The processing component 32 is used to: construct a temporary dedicated satellite communication link according to the emergency location information and preliminary medical condition description in the received emergency medical distress signal, and optimize the parameters of the temporary dedicated satellite communication link to obtain an optimized communication link; use the optimized communication link to combine augmented reality technology and computer vision algorithms to perform three-dimensional reconstruction of the on-site environment, realize high-definition audio and video calls and real-time data sharing between medical experts and on-site rescue personnel, and apply machine learning models to analyze the description of the severity of the disease, predict the development of the disease, and generate disease prediction results; based on the disease prediction results, a reinforcement learning algorithm is used to consider weather changes, sea level changes, and other factors. The distribution strategy of special medical supplies is optimized based on the factors of flow direction to obtain the optimal distribution route planning; according to the optimized communication link, vital sign monitoring data and specific key environmental indicators are collected and transmitted back to the medical command center in real time, and deep learning algorithms are applied to perform pattern recognition and trend prediction to generate potential risk warnings and form updated medical needs. The specific key environmental indicators include ambient temperature, humidity, air pressure, wind speed, rainfall and ocean current speed; based on the potential risk warnings and updated medical needs, the optimal distribution route planning is re-evaluated and optimized to generate an updated optimal distribution route planning to ensure that the supplies can be delivered to the emergency site in a timely and accurate manner.

[0315] The processing component 32 may include one or more processors to execute computer instructions to complete all or part of the steps in the above method. Of course, the processing component may also be implemented by one or more application-specific integrated circuits (ASICs), digital signal processors (DSPs), digital signal processing devices (DSPDs), programmable logic devices (PLDs), field programmable gate arrays (FPGAs), controllers, microcontrollers, microprocessors or other electronic components to perform the above method.

[0316] The storage component 31 is configured to store various types of data to support operations at the terminal. The storage component can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic memory, flash memory, magnetic disk or optical disk.

[0317] Of course, the computing device may also include other components, such as input / output interfaces, display components, communication components, etc.

[0318] The input / output interface provides an interface between the processing component and the peripheral interface module, which may be an output device, an input device, etc.

[0319] The communication component is configured to facilitate, among other things, wired or wireless communications between the computing device and other devices.

[0320] Among them, the computing device can be a physical device or an elastic computing host provided by a cloud computing platform, etc. In this case, the computing device can refer to a cloud server, and the above-mentioned processing components, storage components, etc. can be basic server resources rented or purchased from the cloud computing platform.

[0321] The present application also provides a computer storage medium storing a computer program, wherein the computer program can achieve the above-mentioned Figure 1 The illustrated embodiment is a method for transmitting emergency information at sea based on satellite communication.

[0322] 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.

[0323] The device embodiments described above are merely illustrative, wherein the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment. Ordinary technicians in this field can understand and implement it without paying creative labor.

[0324] Through the description of the above implementation methods, those skilled in the art can clearly understand that each implementation method can be implemented by means of software plus a necessary general hardware platform, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solution is essentially or the part that contributes to the prior art can be embodied in the form of a software product, and the computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a disk, an optical disk, etc., including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment or some parts of the embodiments.

[0325] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit it. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. A method for transmitting emergency information at sea based on satellite communication, characterized in that: include: Building a temporary dedicated satellite communication link according to the emergency location information and the preliminary medical condition description in the received emergency medical distress signal, and optimizing the parameters of the temporary dedicated satellite communication link to obtain an optimized communication link; The optimized communication link is used to perform three-dimensional reconstruction of the on-site environment in combination with augmented reality technology and computer vision algorithms, so as to realize high-definition audio and video calls and real-time data sharing between medical experts and on-site rescue personnel, and a machine learning model is used to analyze the description of the severity of the disease, predict the progression of the disease, and generate a disease prediction result; Based on the disease prediction results, the distribution strategy of special medical supplies is optimized by taking into account factors such as weather changes and ocean current direction through a reinforcement learning algorithm to obtain the optimal distribution route planning; According to the optimized communication link, vital sign monitoring data and specific key environmental indicators are collected and transmitted back to the medical command center in real time, and deep learning algorithms are applied to perform pattern recognition and trend prediction to generate potential risk warnings and form updated medical needs. The specific key environmental indicators include ambient temperature, humidity, air pressure, wind speed, rainfall and ocean current speed; Based on the potential risk warning and updated medical needs, the optimal distribution route plan is re-evaluated and optimized, and an updated optimal distribution route plan is generated to ensure that supplies can be delivered to the emergency site in a timely and accurate manner.

2. The method according to claim 1, characterized in that The optimized communication link is used to combine augmented reality technology and computer vision algorithms to perform three-dimensional reconstruction of the on-site environment, realize high-definition audio and video calls and real-time data sharing between medical experts and on-site rescue personnel, and apply machine learning models to analyze the description of the severity of the disease, predict the development of the disease, and generate disease prediction results, including: By using the optimized communication link, a high-quality communication connection is established between the emergency site and the medical experts, ensuring stable, low-latency high-definition audio and video calls. The optimized communication link supports large-bandwidth data transmission, ensures the quality and efficiency of real-time data sharing, and obtains a high-quality communication connection; Based on the high-quality communication connection, combined with augmented reality technology, the emergency scene images and video streams are processed to achieve superimposition of virtual information on the actual scene at the medical expert end to assist remote diagnosis and obtain an augmented reality-assisted remote diagnosis environment; Using computer vision algorithms, multi-angle images and videos transmitted from the scene and processed by augmented reality technology are processed to build an accurate three-dimensional model of the emergency scene, and based on the data obtained in the augmented reality-assisted remote diagnosis environment, an accurate three-dimensional reconstruction of the scene is obtained; According to the optimized communication link and the accurate on-site three-dimensional reconstruction, the vital signs data and specific key environmental indicators transmitted from various on-site monitoring equipment are shared in real time and displayed in the form of charts or numerical values ​​on the expert end interface, so as to facilitate instant assessment of the patient's status, promote collaborative work, and obtain real-time shared vital signs data, wherein the specific key environmental indicators include ambient temperature, humidity, air pressure, wind speed, rainfall, and ocean current speed; Preparing input data required by the machine learning model, organizing and preprocessing the preliminary medical condition description and the vital signs data shared in real time through the communication link, and forming a structured data set suitable for machine learning model analysis based on the real-time shared vital signs data; Applying a machine learning model to perform quantitative evaluation and trend prediction on the structured data set suitable for machine learning model analysis, the machine learning model can quantitatively evaluate the severity of the disease and predict the development trend after learning from a large number of cases, and obtain detailed disease analysis results; A disease prediction report is generated based on the detailed disease analysis results and the precise on-site three-dimensional reconstruction.

3. The method according to claim 2, characterized in that The computer vision algorithm is used to process the multi-angle images and videos transmitted from the scene and processed by augmented reality technology to construct an accurate three-dimensional model of the emergency scene, and based on the data obtained in the augmented reality-assisted remote diagnosis environment, an accurate three-dimensional reconstruction of the scene is obtained, including: Using computer vision algorithms, the received multi-angle images and videos are pre-processed to remove noise, correct distortion, and perform color calibration to obtain high-quality input data; Based on the high-quality input data, using a feature detection algorithm, identifying and extracting stable feature points to obtain key feature points for subsequent matching; Based on the key feature points used for subsequent matching, a feature matching algorithm is used to match the key feature points under different viewing angles, and the matching results are optimized by bundle adjustment to improve the accuracy of three-dimensional reconstruction and generate optimized feature point matching results; The positions of the feature points in the optimized feature point matching results in three-dimensional space are calculated by using the triangulation principle to preliminarily form the spatial structure of the emergency scene and obtain a sparse three-dimensional point cloud; Applying a multi-view stereo vision algorithm to fill the gaps between the sparse three-dimensional point clouds to generate a dense three-dimensional point cloud; A Poisson surface reconstruction algorithm is used to construct a continuous three-dimensional surface from the dense three-dimensional point cloud, and texture information of the original image is mapped onto the surface to obtain an accurate three-dimensional model; The precise three-dimensional model is combined with real-time scene information obtained through augmented reality technology to provide medical experts with a comprehensive and intuitive view of the emergency scene, ensure the accuracy and effectiveness of remote diagnosis, and obtain precise three-dimensional reconstruction of the scene.

4. The method according to claim 1, characterized in that: Based on the disease prediction results, the distribution strategy of special medical supplies is optimized by taking into account factors such as weather changes and ocean current direction through a reinforcement learning algorithm to obtain the optimal distribution route planning, including: Based on the disease prediction results, the comprehensive environmental factor assessment module is used to obtain and analyze key environmental parameters of weather changes and ocean current directions in real time to obtain a report on environmental conditions that affect delivery route planning; Based on the environmental condition report, a reinforcement learning algorithm is used to create a multi-objective optimization problem model including transportation cost, time efficiency, and safety factor to obtain an initial distribution strategy; A dynamic adjustment mechanism is introduced into the initial distribution strategy, allowing the distribution strategy to be updated in real time with the latest weather changes and ocean current directions, ensuring that the effectiveness and accuracy of distribution can be maintained even in complex and changing environments, and generating a continuously optimized distribution path plan; By utilizing the interactive learning process between the reinforcement learning model and the actual delivery situation, the delivery route planning is automatically adjusted, and the optimal delivery route planning is obtained based on the latest environmental changes and delivery feedback.

5. The method according to claim 4, characterized in that Based on the environmental condition report, a reinforcement learning algorithm is applied to create a multi-objective optimization problem model including transportation cost, time efficiency, and safety factor to obtain an initial distribution strategy, including: Using the environmental condition report, the key environmental parameters of weather changes and ocean current directions are deeply analyzed and processed to obtain the analysis results of factors affecting the distribution route selection and material delivery time; Based on the factor analysis results, multiple optimization objectives are defined, including minimizing transportation costs, maximizing time efficiency, and ensuring the highest safety factor, to generate a multi-objective optimization framework; Applying a reinforcement learning algorithm, based on the multi-objective optimization framework, constructing a multi-objective optimization problem model capable of handling the trade-off relationship between different optimization objectives, and using the objective definition and weight allocation in the multi-objective optimization framework as the basis for model construction; Constructing a simulation environment to simulate different delivery scenarios, providing a training platform for the multi-objective optimization problem model, and applying the multi-objective optimization model to the simulation environment; Through a large number of simulation experiments, the reinforcement learning model is allowed to continuously try to select the delivery path in the simulation environment, and the trade-off relationship between the various objectives is adjusted according to the feedback, gradually converging to the optimal strategy combination, and obtaining a preliminary optimized delivery plan; Based on the preliminary optimized distribution plan, combined with the latest environmental conditions and actual distribution needs, final adjustments are made to generate an initial distribution strategy.

6. The method according to claim 1, characterized in that According to the optimized communication link, vital sign monitoring data and specific key environmental indicators are collected and transmitted back to the medical command center in real time, and deep learning algorithms are applied to perform pattern recognition and trend prediction to generate potential risk warnings and form updated medical needs. The specific key environmental indicators include ambient temperature, humidity, air pressure, wind speed, rainfall and ocean current speed, including: By using the optimized communication link, vital sign monitoring data and specific key environmental indicators at the emergency scene are collected and processed to obtain encoded monitoring data, wherein the specific key environmental indicators include ambient temperature, humidity, air pressure, wind speed, rainfall and ocean current speed; Based on the encoded monitoring data, the encoded monitoring data is processed in real time through the optimized communication link to obtain a data stream that is transmitted to a medical command center in real time; According to the data stream transmitted to the medical command center in real time, applying the deep learning algorithm deployed in the medical command center, analyzing and processing the vital sign monitoring data and specific key environmental indicators to obtain a pattern recognition result, wherein the specific key environmental indicators include ambient temperature, humidity, air pressure, wind speed, rainfall, and ocean current speed; Based on the pattern recognition results, a deep learning model is used to predict the development trend of the patient's condition and generate a trend prediction report; Based on the trend prediction report, the system automatically generates a potential risk warning for each patient, and obtains warning information including current risk factors and future risk warnings; Based on the warning information, the medical command center re-evaluates and adjusts the emergency plan to form updated medical needs.

7. The method according to claim 1, characterized in that Based on the potential risk warning and the updated medical needs, the optimal distribution path plan is re-evaluated and optimized to generate an updated optimal distribution path plan to ensure that the supplies can be delivered to the emergency site in a timely and accurate manner, including: Utilize the potential risk warning to analyze and process the risk factors at the emergency scene and obtain a detailed risk assessment report; Based on the risk assessment report and the updated medical needs, reassess the current medical supplies needs and generate an updated medical supplies needs list; Based on the latest medical supplies demand list, check whether the existing optimal distribution route planning needs to be adjusted to adapt to new demand changes, and preliminarily determine the aspects that need to be optimized; Based on the above-mentioned optimization aspects, the optimal distribution route is replanned taking into account the latest environmental conditions to ensure that the material distribution strategy meets the latest medical needs and risk warnings; Through continuous updating and adjustment, an updated optimal distribution route plan is eventually generated to ensure that special medical supplies can be delivered to the emergency site in a timely and accurate manner.

8. A marine emergency information transmission system based on satellite communication, characterized in that: include: Constructing an optimization module, for constructing a temporary dedicated satellite communication link according to the emergency location information and the preliminary medical condition description in the received emergency medical distress signal, and optimizing the parameters of the temporary dedicated satellite communication link to obtain an optimized communication link; A reconstruction and analysis module is used to utilize the optimized communication link, combined with augmented reality technology and computer vision algorithms, to perform three-dimensional reconstruction of the on-site environment, to achieve high-definition audio and video calls and real-time data sharing between medical experts and on-site rescue personnel, and to apply a machine learning model to analyze the description of the severity of the disease, predict the progression of the disease, and generate a disease prediction result; An optimization processing module is used to optimize the distribution strategy of special medical supplies based on the disease prediction results, by using a reinforcement learning algorithm and taking into account factors such as weather changes and ocean current direction, to obtain an optimal distribution route plan; A collection and formation module is used to collect vital sign monitoring data and specific key environmental indicators according to the optimized communication link, transmit them back to the medical command center in real time, and apply deep learning algorithms to perform pattern recognition and trend prediction, generate potential risk warnings, and form updated medical needs. The specific key environmental indicators include ambient temperature, humidity, air pressure, wind speed, rainfall, and ocean current speed; The evaluation and optimization module is used to re-evaluate and optimize the optimal distribution route plan based on the potential risk warning and updated medical needs, generate an updated optimal distribution route plan, and ensure that the supplies can be delivered to the emergency site in a timely and accurate manner.

9. A computing device, characterized in that It comprises a processing component and a storage component; the storage component stores one or more computer instructions; the one or more computer instructions are used to be called and executed by the processing component to implement a method for transmitting marine emergency information based on satellite communication as described in any one of claims 1 to 7.

10. A computer storage medium, characterized in that: A computer program is stored, and when the computer program is executed by a computer, a method for transmitting emergency information at sea based on satellite communication as claimed in any one of claims 1 to 7 is implemented.

Images