A vehicle-mounted multi-level disaster prevention and early warning method and system based on Beidou communication

Abstract

The present application belongs to the field of vehicle-mounted disaster prevention and early warning technology, and discloses a vehicle-mounted multi-level disaster prevention and early warning method and system based on Beidou communication, the method comprising: obtaining vehicle location information and predicted path information to determine the vehicle driving area; obtaining a detailed geographic information system data packet from a data service platform according to the vehicle driving area; calculating the probability of occurrence of different types of natural disasters in the vehicle driving area and the degree of impact on vehicle safety according to the detailed geographic information system data packet, the vehicle's current speed and the vehicle location information, and obtaining a risk assessment result; classifying the natural disaster risks into risk levels according to the risk assessment results; determining the warning information presentation method according to the risk level, disaster type and driver status, and generating and presenting the warning information according to the corresponding presentation method; thereby providing more accurate and targeted natural disaster warnings and improving the safety of vehicle driving.

Description

Technical Field

[0001] The present application relates to the field of vehicle-mounted disaster prevention and early warning technology, and specifically to a vehicle-mounted multi-level disaster prevention and early warning method and system based on Beidou communication. Background Art

[0002] When a vehicle is driving on the road, its onboard system typically relies on the Beidou satellite navigation system to obtain the vehicle's real-time location, speed, and heading information, and combines this with basic geographic information system (GIS) data stored onboard for positioning and navigation. However, in areas prone to natural disasters, basic GIS data alone is insufficient to address potential risks. Therefore, the system typically requires more detailed GIS data, including high-precision topography, detailed geological structures, detailed water system distribution, and historical records of disaster sites. In addition, real-time environmental monitoring data is crucial for natural disaster warnings. This data comes from meteorological, water conservancy, and transportation departments, as well as IoT sensor networks, providing key monitoring data such as real-time rainfall intensity, river water levels, road water depth, slope soil moisture content, and mountain displacement.

[0003] In-vehicle systems rely on this data when conducting risk assessments. The goal of risk assessment is to determine the type and severity of natural disaster risks that may exist at the vehicle's current location and along its path. This assessment requires comprehensive consideration of multiple factors. However, the overlapping nature of these factors and risk types makes accurate risk assessments significantly challenging.

[0004] After the risk assessment is complete, an early warning is issued based on the results. These warnings are typically presented to the driver via an in-vehicle terminal, and the presentation format can significantly impact driver acceptance and response efficiency. If a disruptive warning message suddenly pops up while the driver is performing a high-attention driving maneuver, it could distract the driver and increase driving risk. Furthermore, existing systems lack the ability to detect whether the driver has received, understood, and processed the warning message, nor do they have a mechanism to further increase the warning intensity or modify the warning method if the driver fails to respond or responds inappropriately.

[0005] In view of the above problems, the existing technology is in urgent need of improvement. Summary of the Invention

[0006] The purpose of this application is to provide a vehicle-mounted multi-level disaster prevention and warning method and system based on Beidou communication, which can provide more accurate and targeted natural disaster warnings and improve the safety of vehicle driving.

[0007] In a first aspect, the present application provides a vehicle-mounted multi-level disaster prevention and warning method based on Beidou communication, which is used to warn of natural disaster risks while the vehicle is driving. The steps of the method include:

[0008] A1. Obtain vehicle location information and predicted path information to determine the vehicle's travel area;

[0009] A2. According to the vehicle travel area, obtain detailed geographic information system data from the data service platform; the detailed geographic information system data package includes meteorological information, hydrological information and geological information of the vehicle travel area;

[0010] A3. Calculate the probability of different types of natural disasters occurring within the vehicle's travel area and their impact on vehicle safety based on the detailed GIS data package, the vehicle's current speed, and its location information, and obtain a risk assessment result;

[0011] A4. Based on the risk assessment results, classify the natural disaster risks into risk levels;

[0012] A5. Determine the presentation method for warning information based on the risk level, disaster type, and driver status, and generate and present warning information according to the corresponding presentation method.

[0013] Preferably, step A1 includes:

[0014] A101. Obtain vehicle location information through the vehicle-mounted Beidou positioning system;

[0015] A102. Obtain destination information, current navigation route information, and historical driving route data from the vehicle navigation system;

[0016] A103. Based on the destination information, current navigation path information, historical route data and real-time vehicle route information, the vehicle's future route is predicted to obtain predicted route information; the real-time route information is a route connected by the vehicle's position within a preset time window before the current moment;

[0017] A104. Use the buffer zone analysis method to construct an initial buffer zone centered on the vehicle's current location and extending along the predicted path.

[0018] A105. Calculate the buffer radius increment based on the road type to adjust the size of the initial buffer zone and obtain the vehicle driving area.

[0019] Preferably, step A103 includes:

[0020] When the destination information and the current navigation path information are both not empty, the Kalman filter algorithm is used to fuse the current navigation path information and the real-time route information to predict the vehicle's future driving route and obtain the predicted path information;

[0021] When both the destination information and the current navigation path information are empty, the Markov model is used to predict the future vehicle driving trajectory based on the historical driving route data, and the predicted trajectory is used as the predicted path information.

[0022] Preferably, A3 includes:

[0023] A301. Obtain a pre-built disaster type association matrix; the disaster type association matrix represents the association probability between different types of natural disasters;

[0024] A302. Extract meteorological information, hydrological information, and geological information within the vehicle's travel area from the detailed geographic information system data packet, and calculate the initial probability of occurrence of various natural disasters based on historical disaster data;

[0025] A303. Based on the disaster type association matrix, the initial probability of occurrence of each type of natural disaster is modified by a weighted average algorithm to obtain a modified probability of occurrence of the disaster;

[0026] A304. Calculate the distance between the vehicle and various risk sources based on vehicle location information. Combined with road type and weather information, use fuzzy logic algorithms to determine the impact of various natural disasters on vehicle safety.

[0027] A305. Based on the associated corrected probability of disaster occurrence, the impact factor, and the current speed of the vehicle, a weighted summation algorithm is used to calculate the comprehensive risk value of each type of natural disaster to the vehicle as the risk assessment result.

[0028] Preferably, step A302 includes:

[0029] Extracting meteorological information, hydrological information, and geological information within the vehicle's travel area from the detailed geographic information system data packet; the meteorological information includes rainfall, wind speed, temperature, and humidity; the hydrological information includes water level, flow, flow velocity, and water quality; and the geological information includes geological structure, rock and soil type, slope, and land cover information;

[0030] Based on the extracted meteorological information, hydrological information and geological information, combined with historical disaster data, a Bayesian network model is used to calculate the initial probability of occurrence of landslides, floods, mudslides and waterlogging in the vehicle's driving area; the historical disaster data includes the disaster type, occurrence time, geographical location, impact range, degree of loss and corresponding meteorological information, hydrological information and geological information.

[0031] Preferably, step A303 includes:

[0032] For each natural disaster, the impact weight of other disaster types on the probability of occurrence of this natural disaster is calculated based on the disaster type association matrix and the preset association probability threshold;

[0033] According to the initial occurrence probability and impact weight of various natural disasters, the occurrence probability of various natural disasters is corrected by the weighted average algorithm to obtain the associated corrected disaster occurrence probability.

[0034] Preferably, step A305 includes:

[0035] Determining the weight coefficients of various natural disasters based on the probability of occurrence of the disaster after the correlation correction;

[0036] According to the current speed of the vehicle, query the speed-influence factor adjustment coefficient table to obtain the influence factor adjustment coefficient corresponding to the current speed of the vehicle;

[0037] Calculate the adjusted impact factors of various natural disasters using multiplication operation based on the impact factors and the impact factor adjustment coefficients;

[0038] According to the weight coefficient corresponding to the associated corrected probability of disaster occurrence and the adjusted impact factor, a weighted summation algorithm is used to calculate the comprehensive risk value of various natural disasters to vehicles as a risk assessment result.

[0039] Preferably, step A5 includes:

[0040] A501. Obtain driver facial images and eye movement data, as well as vehicle operation data, to assess driver status; driver status includes fatigue level and degree of distraction;

[0041] A502. Determine the basic warning intensity and presentation method combination based on the risk level; the presentation method combination includes at least one of the following: visual, auditory, and tactile methods;

[0042] A503. According to the driver's status, query the preset driver status - warning intensity adjustment coefficient table to obtain the corresponding warning intensity adjustment coefficient to adjust the basic warning intensity to obtain the adjusted warning intensity;

[0043] A504. Generate warning information based on the adjusted warning intensity, the combination of presentation methods, and the disaster type, and present the warning information in a corresponding presentation method.

[0044] Preferably, after step A504, the method further includes the following steps:

[0045] A505. Monitor the driver's eye movement data and vehicle operation data to determine whether the driver responds to the warning information. If no response is detected within a preset time, increase the warning information presentation intensity and / or switch the presentation mode until the driver responds.

[0046] In a second aspect, the present application provides a vehicle-mounted multi-level disaster prevention and warning system based on Beidou communication, which is used to provide early warning of natural disaster risks while the vehicle is driving. The system includes:

[0047] The area determination module is used to obtain vehicle location information and predicted path information to determine the vehicle's driving area;

[0048] A data acquisition module is used to obtain a detailed geographic information system data package from a data service platform according to the vehicle's travel area; the detailed geographic information system data package includes meteorological information, hydrological information and geological information of the vehicle's travel area;

[0049] A risk assessment module is used to calculate the probability of occurrence of different types of natural disasters in the vehicle's driving area and the degree of impact on vehicle safety based on the detailed geographic information system data package, the vehicle's current speed and vehicle location information, and obtain a risk assessment result;

[0050] The risk classification module is used to classify the natural disaster risks into risk levels according to the risk assessment results;

[0051] The warning module is used to determine the presentation method of warning information according to the risk level, disaster type and driver status, and generate and present warning information according to the corresponding presentation method.

[0052] Beneficial effects: The present application provides a vehicle-mounted multi-level disaster prevention and warning method and system based on Beidou communication. By acquiring detailed GIS data, conducting risk assessment and multi-level risk classification, compared with the existing technology that only relies on basic GIS data and simple warnings, it can provide more accurate and targeted natural disaster warnings, thereby improving the safety of vehicle driving. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] Figure 1 A flowchart of a vehicle-mounted multi-level disaster prevention and warning method based on Beidou communication provided in an embodiment of the present application.

[0054] Figure 2 A schematic structural diagram of a vehicle-mounted multi-level disaster prevention and warning system based on Beidou communication provided in an embodiment of the present application.

[0055] Explanation of numbers: 1. Area determination module; 2. Data acquisition module; 3. Risk assessment module; 4. Risk classification module; 5. Early warning module. DETAILED DESCRIPTION

[0056] The technical solutions in this application will be clearly and completely described below in conjunction with the drawings in this application. Obviously, the described embodiments are only a part of the embodiments of this application, rather than all of the embodiments. The components of the present application generally described and shown in the drawings here can be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of the present application provided in the drawings is not intended to limit the scope of the application for protection, but merely represents selected embodiments of the present application. Based on the embodiments of the present application, all other embodiments obtained by those skilled in the art without making creative work fall within the scope of protection of this application.

[0057] It should be noted that similar reference numerals and letters represent similar items in the following drawings. Therefore, once an item is defined in one drawing, it does not need to be further defined or explained in subsequent drawings. At the same time, in the description of this application, the terms "first", "second", etc. are only used to distinguish the description and should not be understood as indicating or implying relative importance.

[0058] refer to Figure 1 This application proposes a vehicle-mounted multi-level disaster prevention and early warning method based on Beidou communication, which is used to warn of natural disaster risks while the vehicle is driving. The steps of this method include:

[0059] A1. Obtain vehicle location information and predicted path information to determine the vehicle's travel area;

[0060] A2. Obtain a detailed geographic information system data package (detailed GIS data package) from the data service platform based on the vehicle's travel area. The detailed geographic information system data package includes meteorological, hydrological, and geological information of the vehicle's travel area.

[0061] A3. Calculate the probability of different types of natural disasters occurring within the vehicle's travel area and their impact on vehicle safety based on the detailed GIS data package, the vehicle's current speed, and its location information, and obtain a risk assessment result;

[0062] A4. Based on the risk assessment results, classify the natural disaster risks into risk levels;

[0063] A5. Determine the presentation method for warning information based on the risk level, disaster type, and driver status, and generate and present warning information according to the corresponding presentation method.

[0064] In step A1, a geographic range related to the vehicle's future travel is defined by obtaining the vehicle's current geographic location and predicting its future trajectory. Vehicle location information can be provided by a satellite positioning system, an inertial navigation system, or the vehicle's odometer. Predicted path information can be derived from the vehicle's planned route, historical driving data, or real-time trajectory analysis. The vehicle's travel area can be determined using methods such as buffer zone analysis, based on the vehicle's location or predicted path. This feature is used to limit the scope of subsequent data acquisition and risk assessment, improving processing efficiency.

[0065] In step A2, after determining the area where the vehicle may be traveling, the system retrieves geographic data from external data sources, including meteorological, hydrological, and geological information within that area. This data is more detailed and comprehensive than basic geographic information and can include, for example, high-precision terrain data, detailed geological structure information, real-time rainfall data, and river water level data. This feature provides environmental context for subsequent risk assessments.

[0066] In step A3, detailed environmental data (detailed GIS data packages) and the vehicle's real-time status information (current vehicle speed and location) are comprehensively utilized to calculate the likelihood of various natural disasters occurring within the vehicle's travel area and the potential impact of these disasters on vehicle safety. This calculation method can employ a probabilistic model, fuzzy logic algorithm, or weighted summation algorithm. This feature is used to quantify potential natural disaster risk.

[0067] In step A4, the quantified risk assessment results are mapped to several discrete levels, such as low risk, medium risk, and high risk. The risk level classification results can be obtained by querying a risk assessment result-risk level mapping table, which records the risk levels corresponding to different risk assessment result ranges. This feature is used to convert risk assessment results into identifiers.

[0068] In step A5, the warning signal format (e.g., a combination of visual, auditory, and tactile) and intensity are selected based on the assessed risk level, specific disaster type, and the driver's current physiological or behavioral state. Warning content is generated and issued to the driver. Driver status can include fatigue and concentration. This feature ensures that warning information is effectively communicated to the driver and takes into account the driver's ability to receive it.

[0069] The core innovation of this application lies in that by combining the acquisition of detailed environmental data based on the predicted path with risk assessment combined with the vehicle's own status, and dynamically adjusting the presentation of warning information according to the risk level, disaster type and driver status, it solves the problems existing in the existing technology in data acquisition, risk assessment and warning information communication, and realizes on-board disaster prevention warning.

[0070] Specifically, the method of the present application first obtains the vehicle's current geographic location information and predicts the vehicle's future travel path. Based on this information, the system determines a geographic area within which the vehicle may travel. Next, based on the determined travel area, the system retrieves a detailed geographic information system (GIS) data package containing meteorological, hydrological, and geological information from an external data service platform. Then, using this detailed GIS data package, the vehicle's current speed, and location information, the system calculates the likelihood of different types of natural disasters occurring within the vehicle's travel area and the potential impact of these disasters on vehicle safety, generating a risk assessment result. After obtaining the risk assessment results, the system categorizes the natural disaster risk based on these results. Finally, based on the resulting risk level, the specific disaster type, and the driver's current state, the system determines the presentation method for warning information. The selected method is then generated and issued to the driver. This method, through the sequential execution of these steps, implements a complete process from environmental data acquisition to risk assessment to warning information output. By integrating the vehicle's own state with the driver's, the system improves the targeted and adaptable nature of the warning.

[0071] Through the above scheme, this application determines the driving area by obtaining the vehicle position and predicted path, obtains detailed environmental data in a targeted manner, conducts risk assessment and classification based on the vehicle's own status, and dynamically adjusts the warning presentation method according to the risk level, disaster type and driver status, thereby achieving accurate assessment, timely warning and effective communication of natural disaster risks.

[0072] In some embodiments, step A1 comprises:

[0073] A101. Obtain vehicle location information through the vehicle-mounted Beidou positioning system;

[0074] A102. Obtain destination information, current navigation route information, and historical driving route data from the vehicle navigation system;

[0075] A103. Based on the destination information, current navigation path information, historical route data and real-time vehicle route information, the vehicle's future route is predicted to obtain predicted route information; the real-time route information is a route connected by the vehicle's position within a preset time window before the current moment;

[0076] A104. Use the buffer zone analysis method to construct an initial buffer zone centered on the vehicle's current location and extending along the predicted path.

[0077] A105. Calculate the buffer radius increment based on the road type to adjust the size of the initial buffer zone and obtain the vehicle driving area.

[0078] Among them, predicting the vehicle's future driving route means inferring the geographical path that the vehicle may pass through in the future based on the existing vehicle driving related information. It can be achieved by using technologies such as historical data statistics, navigation planning information correction, or machine learning model prediction.

[0079] Among them, the buffer zone analysis method refers to a method in a geographic information system that uses geographic elements (such as points, lines, and surfaces) as the basis and expands to a certain distance around them to form an area with a specific width.

[0080] The buffer radius increment refers to the adjustment value of the basic buffer distance used in the buffer analysis based on specific geographical or environmental factors (such as road type) to change the size of the final buffer area.

[0081] The method of this application, through the synergistic effect of the aforementioned steps, implements the following operating principles: First, the vehicle's precise location is obtained through the onboard Beidou positioning system, providing an accurate geographic starting point for subsequent route prediction and area determination. Next, the system obtains multi-source information from the vehicle's navigation system, including the driver's planned destination, the current navigation route, and the vehicle's historical driving route data. This information, combined with the vehicle's recent real-time driving trajectory, serves as the basis for predicting the vehicle's future driving route. Based on this multi-source information, the system is able to predict a future route that better aligns with the vehicle's actual driving intentions and habits. Subsequently, utilizing buffer analysis techniques within a geographic information system, starting from the vehicle's current location and extending along the predicted future route, an initial buffer zone is constructed that initially covers the vehicle's possible travel range. To ensure a more accurate and reasonable determination of the area, the system further considers the type of road the vehicle is currently on. Different road types correspond to different driving speeds, safety margins, and potential risk distribution characteristics. Therefore, the system calculates an adjustment to the buffer radius based on the road type and uses this adjustment to modify the size of the initial buffer zone. For example, a larger buffer zone may be required on highways to cover longer prediction distances and potential risk ranges, while a smaller buffer zone may be required on urban roads to focus on closer areas. Through this dynamic adjustment based on road type, the final vehicle travel area can more accurately reflect the actual geographic range of risks the vehicle may face, avoiding the problem of a range that is too large or too small. This accurate and reasonable area determination provides a reliable foundation for the subsequent acquisition of detailed geographic information system data packages within the area from the data service platform, ensuring that the acquired data effectively covers the potential risk area while reducing the acquisition and processing of redundant data, thereby improving the efficiency and accuracy of the entire early warning system.

[0082] The method of the present application is described below with reference to a specific embodiment. In a specific vehicle-mounted system, vehicle location information is acquired via a positioning module that supports Beidou signal reception. The vehicle navigation system can provide a user-set destination, the current navigation route plan, and historical driving trajectory data stored in local memory. Real-time route information can be obtained by continuously recording location points output by the positioning module and connecting them within a preset time window to form trajectory segments. When predicting the vehicle's future driving route, if the navigation system is operational and has a clear destination and route, the system may prioritize the navigation route and, combined with the real-time route information, modify the navigation route using a fusion algorithm (e.g., a weighted averaging or filtering method) to obtain a predicted route. If the navigation system is inoperative or has no clear destination, the system may analyze the historical driving route data to identify the most likely driving pattern at the vehicle's current location and direction, and based on this, predict future driving trajectory segments as the predicted route. When determining the vehicle's driving area, the system invokes the buffer analysis function in the onboard GIS software library to generate an initial strip buffer along the predicted route, starting from the vehicle's current location and with a preset initial width. The system reads the current road type from the navigation map data, identifying it as a highway, for example. Based on pre-set rules or a lookup table, highways are assigned a larger buffer radius increment. The system applies this increment to the initial buffer width, for example, increasing the initial width by a specific value to create a wider driving area. For other road types, the initial buffer width is adjusted similarly to highways, but the buffer radius increment differs and may be negative (depending on the initial width setting).

[0083] Through the above technical solutions, the present application can achieve the following technical effects: the vehicle location information is obtained through the vehicle-mounted Beidou positioning system, thereby improving the accuracy of the location information. The future driving route of the vehicle is predicted by comprehensively utilizing the destination information, current navigation path information, historical driving route data and the vehicle's real-time route information, thereby improving the reliability of the predicted route. By adopting the buffer analysis method and adjusting the buffer size in combination with the road type, a vehicle driving area that is more in line with the actual driving conditions is determined, avoiding unreasonable regional scope. This enables the detailed geographic information system data packets subsequently obtained to effectively cover areas where the vehicle may be at risk, while reducing the acquisition and processing of redundant data, and improving the efficiency and accuracy of subsequent data acquisition and risk assessment.

[0084] Preferably, step A103 may include:

[0085] When the destination information and the current navigation path information are both not empty, the Kalman filter algorithm is used to fuse the current navigation path information and the real-time route information to predict the vehicle's future driving route and obtain the predicted path information;

[0086] When both the destination information and the current navigation path information are empty, the Markov model is used to predict the future vehicle driving trajectory based on the historical driving route data, and the predicted trajectory is used as the predicted path information.

[0087] This prediction method incorporates the Kalman filter algorithm and the Markov model. The Kalman filter is a recursive algorithm for optimally estimating noisy dynamic systems. It can fuse data from different sources to predict and update the system's state. The Markov model is a statistical model that describes the transition probabilities between a series of possible states and can learn and predict future state sequences based on historical data.

[0088] This solution aims to predict the future driving route of a vehicle. Different prediction algorithms are used according to the two different driving scenarios: whether there is destination information or current navigation path information, to improve the accuracy and adaptability of the prediction.

[0089] Specifically, when both the destination information and the current navigation path information are not empty, this indicates that the vehicle is using the navigation system and has a clear driving goal and planned path. In this scenario, the vehicle's future driving route will largely follow the navigation path, but actual driving may include positioning errors or driver adjustments. Therefore, by fusing the current navigation path information (as the desired trajectory) with real-time route information (reflecting actual driving conditions) through the Kalman filter algorithm, this method leverages the Kalman filter's advantage in optimally estimating noisy dynamic systems to more accurately predict the vehicle's future trajectory under navigation guidance. Fusion of the navigation path information helps correct for potential biases that might arise from predictions based solely on the real-time route, while fusion of the real-time route information reflects the driver's actual execution of the navigation path, making the predictions more realistic. The Kalman filter algorithm uses the current navigation path information as the system's desired state or reference input and real-time position and motion information as observations. Through prediction and update steps, it estimates the optimal sequence of vehicle positions over a period of time, forming a predicted path.

[0090] When both the destination information and the current navigation route are empty, this indicates that the vehicle is not using the navigation system, and the driver may be driving freely. In this scenario, the vehicle's future driving behavior is more dependent on the driver's habits and historical behavior patterns. Therefore, a Markov model can be used to learn the driver's driving habits and route selection preferences under different locations and road conditions based on historical route data. Using this Markov model, the vehicle's most likely future driving trajectory, without navigation constraints, can be predicted based on the vehicle's current real-time route information. This approach captures the driver's individual driving style and preference for familiar routes, filling the gap in prediction when navigation information is unavailable. Specifically, the vehicle system stores a database of historical routes. This database records the vehicle's trajectories at different times and along different road sections. The Markov model can be pre-trained based on this historical data to learn the transition probabilities between different road sections or locations. During prediction, the vehicle's current real-time location and recent driving trajectory (real-time route information) are used as the current state. The trained Markov model is then used to predict the most likely series of road sections or locations to be traveled in the future, thereby generating predicted route information.

[0091] By distinguishing between scenarios with and without navigation information and employing two different prediction algorithms, the Kalman filter and the Markov model, tailored to each scenario, this solution can more accurately and robustly predict the vehicle's future path under various driving conditions. This resulting predicted path information is more reliable, providing a more robust foundation for subsequent determination of vehicle travel areas, acquisition of detailed geographic information system data packages, risk assessment, and generation of warning information, thereby enhancing the effectiveness of the entire disaster prevention and early warning approach.

[0092] In some embodiments, the A3 comprises:

[0093] A301. Obtain a pre-built disaster type association matrix; the disaster type association matrix represents the association probability between different types of natural disasters;

[0094] A302. Extract meteorological information, hydrological information, and geological information within the vehicle's travel area from the detailed geographic information system data packet, and calculate the initial probability of occurrence of various natural disasters based on historical disaster data;

[0095] A303. Based on the disaster type association matrix, the initial probability of occurrence of each type of natural disaster is modified by a weighted average algorithm to obtain a modified probability of occurrence of the disaster;

[0096] A304. Calculate the distance between the vehicle and various risk sources based on vehicle location information. Combined with road type and weather information, use fuzzy logic algorithms to determine the impact of various natural disasters on vehicle safety.

[0097] A305. Based on the associated corrected probability of disaster occurrence, the impact factor, and the current speed of the vehicle, a weighted summation algorithm is used to calculate the comprehensive risk value of each type of natural disaster to the vehicle as the risk assessment result.

[0098] The disaster type association matrix is ​​a two-dimensional data structure used to quantify the probability of mutual influence between different types of natural disasters (i.e., the probability that one natural disaster will cause another). This can be achieved using a matrix constructed based on statistical analysis of historical disaster data or expert experience. The rows and columns of the matrix represent different natural disaster types, and the elements of the matrix represent the probability of mutual influence between the corresponding two natural disasters.

[0099] Among them, the weighted average algorithm refers to a method of calculating an average value, in which each value is multiplied by a weight factor, and then the results are added and divided by the sum of the weights. It can be achieved by assigning weights according to the importance of different factors.

[0100] Among them, fuzzy logic algorithm refers to a method of reasoning and decision-making based on fuzzy set theory, which can handle imprecise or uncertain information. It can be implemented by building a fuzzy rule base and membership function.

[0101] The weighted summation algorithm refers to a calculation method that multiplies multiple values ​​by corresponding weights and then adds them together to obtain the total. It can be implemented by allocating weights according to the degree of influence of different input items on the result.

[0102] This solution refines the risk assessment process, aiming to address the challenges of data uncertainty, the overlapping of multiple risks, and the integration of multiple factors for accurate assessment. By introducing a hazard type correlation matrix, comprehensively considering various environmental and vehicle status information, and employing corresponding algorithms, the accuracy and reliability of risk assessment are improved. Specifically, obtaining a pre-constructed hazard type correlation matrix provides foundational data for subsequent consideration of the interplay and associated probabilities between different natural hazard types, helping to address the issue of assessing the overlapping nature of multiple risk types. Meteorological, hydrological, and geological information within the vehicle's travel area is extracted from detailed geographic information system data packages. Based on historical disaster data, the initial probability of occurrence of each natural hazard is calculated. This forms the foundation for risk assessment, leveraging existing environmental information and historical experience to provide a preliminary assessment of the likelihood of each hazard. Based on the hazard type correlation matrix, the calculated initial probability of occurrence of each natural hazard is corrected using a weighted average algorithm to produce a corrected probability of occurrence. This step accounts for the interdependence between different hazards. For example, certain meteorological conditions may increase the risk of both landslides and debris flows. Correction through the correlation matrix ensures that the probability assessment is more realistic and reflects the impact of overlapping risks. Based on vehicle location information, the distance between the vehicle and various risk sources is calculated. Fuzzy logic algorithms are then used to determine the impact factors of various natural disasters on vehicle safety, combined with road type and weather conditions. This step assesses the impact of a disaster on vehicle safety. By considering multiple factors such as the distance between the vehicle and the risk source, road type, and weather conditions, and using fuzzy logic algorithms to process these potentially uncertain inputs, the potential impact of a disaster on the vehicle can be more accurately quantified, rather than simply determining whether a disaster will occur. Based on the correlated and corrected probability of the disaster, the calculated impact factors, and the vehicle's current speed, a weighted summation algorithm is used to calculate a comprehensive risk value for each natural disaster to the vehicle, serving as the risk assessment result. This step integrates the likelihood of a disaster occurring, the potential impact on the vehicle, and the vehicle's own state into the assessment. The faster the vehicle's speed, the shorter the driver's reaction and avoidance time, and the higher the risk. This weighted summation yields a comprehensive risk value that comprehensively reflects the threat level of various natural disasters to vehicle safety in the current scenario, providing a quantitative basis for subsequent risk stratification. This solution is based on determining the vehicle's driving area and obtaining detailed geographic information system data packages, so that risk assessment can be targeted at the areas and environmental conditions that the vehicle may actually face, thereby improving the targeted nature of the assessment.

[0103] Through the above-mentioned technical means, this solution can more accurately calculate the probability of different types of natural disasters occurring in the vehicle's driving area and the degree of impact on vehicle safety. It overcomes the challenges brought about by input data uncertainty, the superposition of multiple risk types, and insufficient consideration of comprehensive factors, improves the accuracy and reliability of risk assessment, and provides a more solid foundation for subsequent risk warnings.

[0104] Preferably, step A302 may include:

[0105] Extracting meteorological information, hydrological information, and geological information within the vehicle's travel area from the detailed geographic information system data packet; the meteorological information includes rainfall, wind speed, temperature, and humidity; the hydrological information includes water level, flow, flow velocity, and water quality; and the geological information includes geological structure, rock and soil type, slope, and land cover information;

[0106] Based on the extracted meteorological information, hydrological information and geological information, combined with historical disaster data, a Bayesian network model is used to calculate the initial probability of occurrence of landslides, floods, mudslides and waterlogging in the vehicle's driving area; the historical disaster data includes the disaster type, occurrence time, geographical location, impact range, degree of loss and corresponding meteorological information, hydrological information and geological information.

[0107] Among them, the detailed geographic information system data package can be stored in the format of Shapefile, GeoJSON or GeoPackage.

[0108] Among them, the Bayesian network model refers to a directed acyclic graph model, in which nodes represent random variables and edges represent conditional dependencies between variables. The intensity of this dependency is described by conditional probability distribution, and is used for probabilistic reasoning and uncertainty modeling. It can use structural learning algorithms to determine the network structure and parameter learning algorithms to determine the conditional probability distribution.

[0109] This application extracts meteorological, hydrological, and geological information within the vehicle's travel area from a detailed geographic information system data package, using this information as key input for calculating the initial probability of occurrence. Meteorological, hydrological, and geological information are environmental factors closely related to the occurrence of natural disasters such as landslides, floods, debris flows, and waterlogging. Combined with historical disaster data, this data contains the environmental conditions and disaster characteristics of past disasters, providing a basis for understanding the relationship between environmental factors and disaster occurrence. Using a Bayesian network model, a probabilistic model is constructed between environmental factors and the occurrence of specific disasters. The Bayesian network model can effectively represent and handle complex dependencies between variables and data uncertainty. By using the extracted current environmental information as evidence input into the Bayesian network, the model can perform probabilistic reasoning and calculate the initial probability of occurrence of four specific disasters: landslides, floods, debris flows, and waterlogging, within the vehicle's travel area under current environmental conditions. The Bayesian network model's ability to handle uncertainty enables reasonable probabilistic inference even when input data is missing or noisy, thus overcoming the difficulty of traditional methods in effectively handling uncertain data and complex associations. The calculated initial probability of occurrence serves as the basic input for subsequent risk assessment steps, such as for correlation correction and comprehensive risk value calculation, which improves the accuracy and reliability of the overall risk assessment.

[0110] In one embodiment, a detailed geographic information system data package uses the GeoPackage format and contains layers such as elevation data, soil type distribution maps, river and lake boundaries, historical landslide locations, and historical flood-inundated areas. Meteorological, hydrological, and geological information are extracted from the corresponding layers or attribute tables in the data package. For example, spatial queries can be used to obtain average rainfall, water levels of major rivers, soil types, and slope information within the vehicle's travel area. Historical disaster data is stored in a relational database, containing detailed records of each historical disaster event. A Bayesian network model is pre-built and stored in the vehicle terminal. Its nodes include variables such as meteorological information (such as rainfall), hydrological information (such as water levels), geological information (such as slope and soil types), and historical disaster events (such as historical landslides, floods, debris flows, and waterlogging). Statistical analysis and machine learning training of historical disaster data are used to determine the network structure and conditional probability table. During operation, the extracted current meteorological, hydrological, and geological information is used as evidence input for the corresponding nodes in the Bayesian network, and the belief propagation algorithm is used to calculate the posterior probability of landslide, flood, debris flow, and waterlogging nodes, that is, the initial occurrence probability.

[0111] By extracting meteorological, hydrological and geological information closely related to specific natural disasters from detailed geographic information system data packages, and combining it with historical disaster data, the Bayesian network model is used to calculate the initial occurrence probability. This can more accurately capture the complex probabilistic relationship between environmental factors and disaster occurrence, effectively handle data uncertainty, and improve the accuracy and robustness of the calculation of the initial occurrence probability of landslides, floods, mudslides and waterlogging in the vehicle's driving area.

[0112] Preferably, step A303 may include:

[0113] For each natural disaster, the impact weight of other disaster types on the probability of occurrence of this natural disaster is calculated based on the disaster type association matrix and the preset association probability threshold;

[0114] According to the initial occurrence probability and impact weight of various natural disasters, the occurrence probability of various natural disasters is corrected by the weighted average algorithm to obtain the associated corrected disaster occurrence probability.

[0115] The association probability threshold is a preset value used to filter out disaster types with weak associations, ensuring that only associated disasters with a significant impact on the probability of the target disaster are considered. The impact weight quantifies the degree to which other disaster types influence the probability of the target disaster. Its calculation method can be determined based on the values ​​in the association probability matrix and the association probability threshold. For example, the impact weight can be simply set to the association probability itself, or the association probability minus the association probability threshold. For other disaster types with an association probability less than the association probability threshold, the corresponding impact weight can be set to 0.

[0116] This solution introduces the concept of impact weights and describes in detail how to calculate and apply these weights, providing a more refined approach to modifying initial occurrence probabilities based on disaster correlations. Specifically, for each natural disaster, the system first consults a pre-built disaster type correlation matrix, which stores the correlation probabilities between different natural disaster types. Based on a preset correlation probability threshold, the system identifies other disaster types with a correlation probability above the threshold with the current target disaster type. Based on the correlation probabilities in the correlation matrix and the set threshold, the system calculates the specific impact weights of these associated disaster types on the probability of the target disaster. This weight quantifies the effect of the associated disaster on the likelihood of the target disaster. After calculating the impact weights of each associated disaster on the target disaster, the system combines these weights with the initial occurrence probabilities of each natural disaster. The initial occurrence probabilities are calculated independently based on historical data and current environmental information, reflecting the likelihood of occurrence without considering the influence of correlations. Using a weighted average algorithm, the initial probabilities are used as a basis and adjusted based on the calculated impact weights. Related disasters with larger weights have a more significant effect on the probability of the target disaster. For example, the correction factor can be calculated by multiplying the weighted sum of the initial probabilities of other disaster types by the sum of their weights. This correction factor can then be added to the initial probability of the target natural disaster to obtain the associated and corrected probability of the target natural disaster. This weighted average approach comprehensively considers the target disaster's own occurrence basis and the impact of other related disasters on its likelihood of occurrence, thereby obtaining a more comprehensive and accurate associated and corrected probability of the disaster.

[0117] Through these technical means, this application can more precisely quantify the specific impact of different related disasters on the probability of a target disaster, making the correction process more accurate. By applying a weighted average correction to the initial probability of occurrence based on the calculated impact weights, it can more comprehensively reflect the likelihood of a specific natural disaster occurring in an environment where multiple disasters may affect each other, thereby improving the accuracy of the final probability assessment.

[0118] Preferably, step A305 may include:

[0119] Determining the weight coefficients of various natural disasters based on the probability of occurrence of the disaster after the correlation correction;

[0120] According to the current speed of the vehicle, query the speed-influence factor adjustment coefficient table to obtain the influence factor adjustment coefficient corresponding to the current speed of the vehicle;

[0121] Calculate the adjusted impact factors of various natural disasters using multiplication operation based on the impact factors and the impact factor adjustment coefficients;

[0122] According to the weight coefficient corresponding to the associated corrected probability of disaster occurrence and the adjusted impact factor, a weighted summation algorithm is used to calculate the comprehensive risk value of various natural disasters to vehicles as a risk assessment result.

[0123] Among them, the weight coefficient refers to a measure of the relative importance of various natural disasters in the calculation of the comprehensive risk value. It can be determined proportionally based on the probability of disaster occurrence after correlation correction. For example, the higher the probability of occurrence, the greater the weight coefficient.

[0124] The speed-impact factor adjustment coefficient table is a pre-set data structure that stores the mapping between different vehicle speed ranges or specific speed values ​​and corresponding impact factor adjustment coefficients. This can be implemented using a table, function, or lookup table. The impact factor adjustment coefficient is a numerical factor used to dynamically adjust the impact of natural disasters on vehicle safety. Its value depends on the vehicle's current speed and reflects the impact of speed changes on driver reaction time, vehicle handling, braking distance, and other factors, thereby changing the actual threat level of the disaster.

[0125] Among them, the adjusted impact factor refers to the value obtained by multiplying the original impact factor by the impact factor adjustment coefficient, which can better reflect the actual impact of natural disasters on vehicle safety at the current speed of the vehicle.

[0126] Among them, the comprehensive risk value refers to the value obtained by weighted summation based on the comprehensive consideration of the probability of occurrence of natural disasters and the adjusted influencing factors. It is used to quantify the overall risk level of various natural disasters to vehicles.

[0127] Based on the above technical features, the comprehensive risk value calculation method of the present application operates as follows: The scheme is based on the correlated and corrected disaster probability and impact factor obtained in the previous steps. First, the correlated and corrected probability is converted into a weight coefficient, giving disasters with a higher probability of occurrence a greater weight in subsequent calculations. Simultaneously, the vehicle's current speed is obtained, and a preset speed-impact factor adjustment coefficient table is used to find the adjustment coefficient corresponding to the current speed. This adjustment coefficient reflects the dynamic impact of speed on the vehicle's disaster response capabilities. Then, the impact factor calculated in the previous step is multiplied by the found adjustment coefficient to obtain the adjusted impact factor. This adjusted impact factor dynamically reflects the actual impact of the disaster on the vehicle at the current speed. Finally, the weight coefficient determined based on the probability of occurrence and the adjusted impact factor are weighted and summed. The weight coefficient corresponding to the probability of occurrence serves as the weight, and the adjusted impact factor serves as the weighted term. This calculation method comprehensively considers the probability of a disaster and its actual impact at the current speed, thereby obtaining a more comprehensive and accurate comprehensive risk value. This solution is combined with the above steps. On the basis of the above steps calculating the probability of occurrence taking into account the correlation and the influencing factors based on factors such as distance, road type, and weather, it further introduces the dynamic influence of vehicle speed and uses the probability of occurrence as the weight for weighting. The final risk assessment result not only takes into account the characteristics and environmental factors of the disaster itself, but also the dynamic state of the vehicle itself, thereby improving the accuracy and practicality of the risk assessment.

[0128] Through the above method, the present application can achieve the following technical effects: by determining the weight coefficient based on the probability of disaster occurrence after correlation correction, the disaster with a higher probability of occurrence has a greater weight in the calculation of the comprehensive risk value, which more reasonably reflects the contribution of the possibility of different disasters to the overall risk. By querying the speed-influence factor adjustment coefficient table according to the current speed of the vehicle to obtain the adjustment coefficient, and using it to adjust the influence factor, the influence factor can dynamically reflect the impact of the vehicle speed on the driver's reaction, vehicle control, etc., thereby more accurately assessing the actual threat level of the disaster under the current driving state. By performing a weighted summation based on the weight based on the probability of occurrence and the influence factor based on the speed adjustment, the comprehensive risk value obtained is more comprehensive, more dynamic, and more accurate, which improves the reliability of the risk assessment results and provides a more solid foundation for subsequent risk level classification and early warning.

[0129] In some embodiments, step A5 comprises:

[0130] A501. Obtain driver facial images and eye movement data, as well as vehicle operation data, to assess driver status; driver status includes fatigue level and degree of distraction;

[0131] A502. Determine the basic warning intensity and presentation method combination based on the risk level; the presentation method combination includes at least one of the following: visual, auditory, and tactile methods;

[0132] A503. According to the driver's status, query the preset driver status - warning intensity adjustment coefficient table to obtain the corresponding warning intensity adjustment coefficient to adjust the basic warning intensity to obtain the adjusted warning intensity;

[0133] A504. Generate warning information based on the adjusted warning intensity, the combination of presentation methods, and the disaster type, and present the warning information in a corresponding presentation method.

[0134] Among them, the driver state refers to the driver's current physiological and psychological state, reflecting his or her ability to perceive and process information, and may include fatigue level, concentration level, etc. The fatigue level refers to the driver's fatigue level, which can be evaluated based on facial features, eye movement data, and vehicle operation data, and can be divided into different levels. The degree of distraction refers to the driver's concentration, which can be evaluated based on eye movement data, facial orientation, and vehicle operation data, and can be quantified into different levels. The fatigue level and distraction level can be identified using their respective pre-trained recognition models, which are trained using historical facial images, historical eye movement data, and historical vehicle operation data as training data.

[0135] Among them, the basic warning intensity refers to the initial intensity level of the warning information determined according to the risk level. The intensity can be reflected in visual brightness, flashing frequency, auditory volume, tone, tactile vibration frequency, amplitude and other aspects.

[0136] The presentation mode combination refers to a combination of sensory channels used to present warning information to the driver, which may include at least one of vision, hearing and touch.

[0137] The preset driver state-warning intensity adjustment coefficient table refers to a lookup table that stores the relationship between different driver states and corresponding warning intensity adjustment coefficients. This table can be pre-set based on experimental data or expert experience. The warning intensity adjustment coefficient is a multiplicative or additive coefficient used to adjust the basic warning intensity. This coefficient reflects the degree to which the warning intensity needs to be strengthened or weakened under a specific driver state. The adjusted warning intensity is the final warning intensity obtained by adjusting the basic warning intensity according to the warning intensity adjustment coefficient.

[0138] Through the above steps, this application implements a technical solution that dynamically adjusts the warning intensity and presentation method based on the driver's real-time status. The system first obtains the driver's real-time status data and assesses their fatigue and distraction. Simultaneously, based on the risk level determined in the above steps, a standard combination of warning intensity and presentation method is determined. Then, using the assessed driver status, an adjustment coefficient is retrieved by querying a preset adjustment table. This adjustment coefficient is used to modify the standard warning intensity, resulting in an adjusted warning intensity tailored to the driver's current state. Finally, combining this adjusted intensity, the predetermined presentation method combination, and the specific disaster type, a warning message is generated and presented to the driver. This entire process forms a closed loop, from sensing the driver's status to dynamically adjusting the warning output, ensuring that the warning information is delivered in a manner that is most likely to be perceived and understood by the driver. By combining the risk level and disaster type determined in the above steps with the driver's real-time status, the warning information issued by the system not only reflects the risks of the external environment but also takes into account the individual driver's receptive ability, thereby improving the relevance and effectiveness of the warning.

[0139] Through the above technical solution, the present application can dynamically adjust the warning intensity and presentation method according to the real-time status of the driver. When the driver is tired or distracted, the warning intensity is enhanced, which increases the possibility of the warning information being perceived and noticed. When the driver is in good condition, the warning intensity can be adjusted appropriately to avoid excessive interference. This adaptive adjustment based on the driver's status enables the warning information to reach the driver more effectively and attract their sufficient attention, thereby improving the effectiveness of the warning and reducing the risk caused by the driver's failure to respond to the warning in a timely manner.

[0140] Preferably, after step A504, the following steps may also be included:

[0141] A505. Monitor the driver's eye movement data and vehicle operation data to determine whether the driver responds to the warning information. If no response is detected within a preset time, increase the warning information presentation intensity and / or switch the presentation mode until the driver responds.

[0142] Eye movement data refers to data reflecting the driver's physiological state, such as eye movement, gaze point, and pupil size. It can be obtained using eye tracking equipment or eye movement analysis algorithms integrated into vehicle cameras. Vehicle operation data refers to data reflecting the driver's control operations on the vehicle. It can include information such as steering angle, accelerator pedal position, brake pedal position, vehicle speed, and gear position. It can usually be obtained through the vehicle's CAN bus system.

[0143] Determining whether the driver has responded to the warning message involves analyzing, based on monitored eye movement data and vehicle operation data, whether the driver has shown attention to the warning message or has taken driving actions related to the warning message. This can use preset rules or machine learning models to identify specific eye movement patterns (such as looking at the warning display) or changes in vehicle operation (such as slowing down or changing lanes). The preset time is a configurable interval (which can be set according to actual needs) that the system waits for the driver to respond after issuing a warning message.

[0144] Increasing the intensity of warning information presentation refers to improving the perception of warning signals, which can be achieved by increasing the warning volume, enhancing the visual flashing frequency or brightness, increasing the tactile vibration amplitude, etc. Switching the presentation mode refers to changing the modal combination of warning information transmission, which can be achieved by switching from a single visual prompt to a combined visual and auditory prompt, or adding tactile prompts, etc.

[0145] This solution goes beyond generating and presenting warnings by further incorporating a driver response monitoring and feedback mechanism. After the system generates and presents warnings based on risk level, hazard type, and driver status, it continuously monitors the driver's eye movement and vehicle operation data. By analyzing this data, the system can determine whether the driver has paid attention to the warning or taken appropriate driving actions within a preset time window. This monitoring and judgment mechanism is critical to ensuring the effectiveness of warnings. If the system fails to detect a driver response within the preset time window, it indicates that the driver may have failed to notice the warning or failed to understand and act in a timely manner. In this case, the system does not terminate the warning, but instead proactively implements enhanced measures. By increasing the intensity of the warning, such as increasing the loudness of the sound, the prominence of the visual signal, or the strength of the tactile feedback, and / or switching the presentation method of the warning, such as adding auditory or tactile cues to a single visual cue, the system can more effectively attract the driver's attention and overcome any potential perceptual impairments or distractions. This dynamic adjustment and enhancement of warnings significantly increases the compulsion and perceptibility of warnings, until the system detects that the driver has taken a relevant response. This feedback loop, which monitors after a warning is issued and dynamically adjusts the warning intensity and method based on the driver's status, enables the warning information to be received and processed more effectively by the driver, thereby prompting the driver to take timely risk avoidance measures and effectively respond to natural disaster risks.

[0146] refer to Figure 2 This application provides a vehicle-mounted multi-level disaster prevention and warning system based on Beidou communication, which is used to warn of natural disaster risks while the vehicle is driving. The system includes:

[0147] Region determination module 1, used to obtain vehicle location information and predicted path information to determine the vehicle's driving area (for details, please refer to step A1 above);

[0148] Data acquisition module 2 is used to obtain a detailed geographic information system data package from the data service platform based on the vehicle's travel area; the detailed geographic information system data package includes meteorological information, hydrological information, and geological information of the vehicle's travel area (for details, please refer to step A2 above);

[0149] Risk assessment module 3 is used to calculate the probability of different types of natural disasters occurring in the vehicle's driving area and the degree of impact on vehicle safety based on the detailed geographic information system data package, the vehicle's current speed, and the vehicle's location information, and obtain a risk assessment result (the specific process can be referred to in step A3 above);

[0150] Risk grading module 4 is used to classify the natural disaster risk into risk levels based on the risk assessment results (the specific process can be found in step A4 above);

[0151] The warning module 5 is used to determine the presentation method of the warning information according to the risk level, disaster type and driver status, and generate and present the warning information according to the corresponding presentation method (the specific process can be referred to step A5 above).

[0152] The above description is merely an embodiment of the present application and is not intended to limit the scope of protection of the present application. For those skilled in the art, various modifications and variations of the present application are possible. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present application shall be included in the scope of protection of the present application.

Claims

1. A vehicle-mounted multi-level disaster prevention and early warning method based on Beidou communication, used to provide early warning of natural disaster risks during vehicle driving, characterized in that: The steps of the method include: A1. Obtain vehicle location information and predicted path information to determine the vehicle's travel area; A2. According to the vehicle travel area, obtain detailed geographic information system data from the data service platform; the detailed geographic information system data package includes meteorological information, hydrological information and geological information of the vehicle travel area; A3. Calculate the probability of different types of natural disasters occurring within the vehicle's travel area and their impact on vehicle safety based on the detailed GIS data package, the vehicle's current speed, and its location information, and obtain a risk assessment result; A4. Based on the risk assessment results, classify the natural disaster risks into risk levels; A5. Determine the presentation method for warning information based on the risk level, disaster type, and driver status, and generate and present warning information accordingly; The A3 includes: A301. Obtain a pre-built disaster type association matrix; the disaster type association matrix represents the association probability between different types of natural disasters; A302. Extract meteorological information, hydrological information, and geological information within the vehicle's travel area from the detailed geographic information system data packet, and calculate the initial probability of occurrence of various natural disasters based on historical disaster data; A303. Based on the disaster type association matrix, the initial probability of occurrence of each type of natural disaster is modified by a weighted average algorithm to obtain a modified probability of occurrence of the disaster; A304. Calculate the distance between the vehicle and various risk sources based on vehicle location information. Combined with road type and weather information, use fuzzy logic algorithms to determine the impact of various natural disasters on vehicle safety. A305. Based on the associated corrected disaster probability, the impact factor, and the current vehicle speed, a weighted summation algorithm is used to calculate the comprehensive risk value of each natural disaster to the vehicle as the risk assessment result; Step A305 includes: Determining the weight coefficients of various natural disasters based on the probability of occurrence of the disaster after the correlation correction; According to the current speed of the vehicle, query the speed-influence factor adjustment coefficient table to obtain the influence factor adjustment coefficient corresponding to the current speed of the vehicle; Calculate the adjusted impact factors of various natural disasters using multiplication operation based on the impact factors and the impact factor adjustment coefficients; According to the weight coefficient corresponding to the associated corrected probability of disaster occurrence and the adjusted impact factor, a weighted summation algorithm is used to calculate the comprehensive risk value of various natural disasters to vehicles as a risk assessment result.

2. The vehicle-mounted multi-level disaster prevention and early warning method based on Beidou communication according to claim 1 is characterized in that: Step A1 includes: A101. Obtain vehicle location information through the vehicle-mounted Beidou positioning system; A102. Obtain destination information, current navigation route information, and historical driving route data from the vehicle navigation system; A103. Based on the destination information, current navigation path information, historical route data and real-time vehicle route information, the vehicle's future route is predicted to obtain predicted route information; the real-time route information is a route connected by the vehicle's position within a preset time window before the current moment; A104. Use the buffer zone analysis method to construct an initial buffer zone centered on the vehicle's current location and extending along the predicted path. A105. Calculate the buffer radius increment based on the road type to adjust the size of the initial buffer zone and obtain the vehicle driving area.

3. The vehicle-mounted multi-level disaster prevention and early warning method based on Beidou communication according to claim 2 is characterized in that: Step A103 includes: When the destination information and the current navigation path information are both not empty, the Kalman filter algorithm is used to fuse the current navigation path information and the real-time route information to predict the vehicle's future driving route and obtain the predicted path information; When both the destination information and the current navigation path information are empty, the Markov model is used to predict the future vehicle driving trajectory based on the historical driving route data, and the predicted trajectory is used as the predicted path information.

4. The vehicle-mounted multi-level disaster prevention and early warning method based on Beidou communication according to claim 1 is characterized in that: Step A302 includes: Extracting meteorological information, hydrological information, and geological information within the vehicle's travel area from the detailed geographic information system data packet; the meteorological information includes rainfall, wind speed, temperature, and humidity; the hydrological information includes water level, flow, flow velocity, and water quality; and the geological information includes geological structure, rock and soil type, slope, and land cover information; Based on the extracted meteorological information, hydrological information and geological information, combined with historical disaster data, a Bayesian network model is used to calculate the initial probability of occurrence of landslides, floods, mudslides and waterlogging in the vehicle's driving area; the historical disaster data includes the disaster type, occurrence time, geographical location, impact range, degree of loss and corresponding meteorological information, hydrological information and geological information.

5. The vehicle-mounted multi-level disaster prevention and early warning method based on Beidou communication according to claim 1 is characterized in that: Step A303 includes: For each natural disaster, the impact weight of other disaster types on the probability of occurrence of this natural disaster is calculated based on the disaster type association matrix and the preset association probability threshold; According to the initial occurrence probability and impact weight of various natural disasters, the occurrence probability of various natural disasters is corrected by the weighted average algorithm to obtain the associated corrected disaster occurrence probability.

6. The vehicle-mounted multi-level disaster prevention and early warning method based on Beidou communication according to claim 1 is characterized in that: Step A5 includes: A501. Obtain driver facial images and eye movement data, as well as vehicle operation data, to assess driver status; driver status includes fatigue level and degree of distraction; A502. Determine the basic warning intensity and presentation method combination based on the risk level; the presentation method combination includes at least one of the following: visual, auditory, and tactile methods; A503. According to the driver's status, query the preset driver status - warning intensity adjustment coefficient table to obtain the corresponding warning intensity adjustment coefficient to adjust the basic warning intensity to obtain the adjusted warning intensity; A504. Generate warning information based on the adjusted warning intensity, the combination of presentation methods, and the disaster type, and present the warning information in a corresponding presentation method.

7. The vehicle-mounted multi-level disaster prevention and early warning method based on Beidou communication according to claim 6, characterized in that: After step A504, the method further includes the following steps: A505. Monitor the driver's eye movement data and vehicle operation data to determine whether the driver responds to the warning information. If no response is detected within a preset time, increase the warning information presentation intensity and / or switch the presentation mode until the driver responds.

8. A vehicle-mounted multi-level disaster prevention and early warning system based on Beidou communication, used to warn of natural disaster risks while the vehicle is traveling, characterized in that: The system includes: The area determination module is used to obtain vehicle location information and predicted path information to determine the vehicle's driving area; A data acquisition module is used to obtain a detailed geographic information system data package from a data service platform according to the vehicle's travel area; the detailed geographic information system data package includes meteorological information, hydrological information and geological information of the vehicle's travel area; A risk assessment module is used to calculate the probability of occurrence of different types of natural disasters in the vehicle's driving area and the degree of impact on vehicle safety based on the detailed geographic information system data package, the vehicle's current speed and vehicle location information, and obtain a risk assessment result; The risk classification module is used to classify the natural disaster risks into risk levels according to the risk assessment results; The warning module is used to determine the presentation method of warning information based on the risk level, disaster type and driver status, and generate and present warning information according to the corresponding presentation method; The risk assessment module calculates the probability of different types of natural disasters occurring in the vehicle's driving area and the degree of impact on vehicle safety based on the detailed geographic information system data packet, the vehicle's current speed, and the vehicle's location information, and when obtaining the risk assessment result, executes: A301. Obtain a pre-built disaster type association matrix; the disaster type association matrix represents the association probability between different types of natural disasters; A302. Extract meteorological information, hydrological information, and geological information within the vehicle's travel area from the detailed geographic information system data packet, and calculate the initial probability of occurrence of various natural disasters based on historical disaster data; A303. Based on the disaster type association matrix, the initial probability of occurrence of each type of natural disaster is modified by a weighted average algorithm to obtain a modified probability of occurrence of the disaster; A304. Calculate the distance between the vehicle and various risk sources based on vehicle location information. Combined with road type and weather information, use fuzzy logic algorithms to determine the impact of various natural disasters on vehicle safety. A305. Based on the associated corrected disaster probability, the impact factor, and the current vehicle speed, a weighted summation algorithm is used to calculate the comprehensive risk value of each natural disaster to the vehicle as the risk assessment result; Step A305 includes: Determining the weight coefficients of various natural disasters based on the probability of occurrence of the disaster after the correlation correction; According to the current speed of the vehicle, query the speed-influence factor adjustment coefficient table to obtain the influence factor adjustment coefficient corresponding to the current speed of the vehicle; Calculate the adjusted impact factors of various natural disasters using multiplication operation based on the impact factors and the impact factor adjustment coefficients; According to the weight coefficient corresponding to the associated corrected probability of disaster occurrence and the adjusted impact factor, a weighted summation algorithm is used to calculate the comprehensive risk value of various natural disasters to vehicles as a risk assessment result.

Images