Risk early warning method, vehicle, storage medium and program product

Abstract

The application discloses a risk early warning method, a vehicle, a storage medium and a program product. The risk early warning method comprises the following steps: in the process of driving the vehicle, a plurality of intersections that the vehicle will pass through are determined according to navigation information; for each intersection, traffic flow data, vehicle flow data and meteorological factors of the vehicle when reaching the intersection are obtained; a feature vector is constructed according to the traffic flow data, the vehicle flow data, the meteorological factors, a turning operation type of the vehicle at the intersection and historical accident data of the intersection; the feature vector is input into a pre-trained prediction model to obtain a risk parameter of the vehicle when reaching the intersection; and a warning operation corresponding to the risk parameter is performed. The scheme provided by the application can identify and warn the risk that the vehicle may encounter in the driving process in advance, so as to ensure driving safety.

Description

Technical Field

[0001] This invention relates to the field of vehicle safety technology, and in particular to risk warning methods, vehicles, storage media, and program products. Background Technology

[0002] With the continuous improvement of vehicle intelligence, vehicle driver assistance functions have become an important technical means to improve driving safety and driving comfort. They can monitor the vehicle's surrounding environment in real time and issue warnings to the driver when there are potential risks, thereby effectively reducing the probability of traffic accidents.

[0003] However, most existing driver assistance systems operate in a passive, reactive mode, only issuing warnings when the vehicle approaches an intersection or blind spot using onboard data collection devices (such as cameras, millimeter-wave radar, and lidar). This mechanism lacks predictive capabilities; if the vehicle is traveling at high speeds, the driver or the autonomous driving system may struggle to take effective hazard avoidance measures in a short time. Furthermore, the data collection devices onboard the vehicle can only collect a limited range of data, leading to inaccurate warnings and a risk of missed or false alarms. Summary of the Invention

[0004] This invention provides a risk warning method, vehicle, storage medium, and program product, which can identify and warn of potential risks that a vehicle may encounter during driving, thereby ensuring driving safety.

[0005] According to one aspect of the present invention, a risk warning method is provided, comprising: during vehicle travel, determining several intersections that the vehicle will pass through based on navigation information; for each intersection, acquiring pedestrian flow data, vehicle flow data, and meteorological factors when the vehicle arrives at the intersection; constructing a feature vector based on the pedestrian flow data, vehicle flow data, meteorological factors, the vehicle's turning operation type at the intersection, and historical accident data of the intersection; inputting the feature vector into a pre-trained prediction model to obtain risk parameters of the vehicle when it arrives at the intersection; and executing a warning operation corresponding to the risk parameters.

[0006] Optionally, the system acquires pedestrian and vehicle traffic data when a vehicle arrives at the intersection, including: acquiring historical pedestrian and vehicle traffic densities for the target area over the past M first time units, where the target area is the area where the vehicle is currently located and includes all intersections the vehicle is about to pass through; inputting the historical pedestrian and vehicle traffic densities into a spatiotemporal graph convolutional network to obtain predicted pedestrian and vehicle traffic densities for the target area over the next N first time units, where M and N are both positive integers and M > N; calculating pedestrian traffic data based on the intersection's latitude and longitude information, the vehicle's arrival time at the intersection, and the predicted pedestrian traffic density; and calculating vehicle traffic data based on the intersection's latitude and longitude information, the vehicle's arrival time at the intersection, and the predicted vehicle traffic density.

[0007] Optionally, obtaining the meteorological factors when the vehicle arrives at the intersection includes: obtaining meteorological forecast data for the intersection in the next P second time units, where P is a positive integer; calculating the perception reliability coefficient corresponding to each second time unit based on the meteorological forecast data for each second time unit; and calculating the meteorological factors based on the time the vehicle arrives at the intersection and the perception reliability coefficient.

[0008] Optionally, meteorological forecast data includes rainfall intensity, visibility, and haze concentration; the perception reliability coefficient is calculated based on the following formula: ;in, Rainfall intensity, The weights corresponding to rainfall intensity For visibility, The weight corresponding to visibility. The concentration of smog, The weights corresponding to the smog concentration.

[0009] Optionally, a feature vector is constructed based on pedestrian flow data, vehicle flow data, meteorological factors, vehicle turning operation types at the intersection, and historical accident data at the intersection. This includes: normalizing the pedestrian flow data, vehicle flow data, and historical accident data at the intersection; determining the turning factor based on the vehicle turning operation types at the intersection; and constructing the feature vector based on the turning factor, meteorological factors, and the normalized pedestrian flow data, vehicle flow data, and historical accident data at the intersection.

[0010] Optionally, after obtaining the risk parameters of the vehicle when it arrives at the intersection, the method further includes: sending a request message to the roadside node of the intersection if the risk parameters are greater than or equal to a preset threshold; receiving real-time environmental data sent by the roadside node based on the request message, and calculating correction parameters based on the real-time environmental data; and correcting the risk parameters based on the correction parameters and the communication delay between the vehicle and the roadside node.

[0011] Optionally, the warning operation corresponding to the risk parameter is executed, including: when the risk parameter is less than the first parameter, recording the vehicle's driving status at the first warning time; when the risk parameter is greater than or equal to the first parameter and less than the second parameter, broadcasting the warning information via voice at the second warning time; when the risk parameter is greater than or equal to the second parameter and less than the third parameter, broadcasting the warning information via voice at the third warning time and displaying the warning location through the central control panel or head-up display system; when the risk parameter is greater than or equal to the third parameter, broadcasting the warning information via voice at the fourth warning time, displaying the warning location through the central control panel or head-up display system, and sending a pre-braking signal to the braking system; wherein, the first warning time is when the straight-line distance between the vehicle's current position and the intersection is equal to the first distance. The warning timing is as follows: the second warning timing is when the straight-line distance between the vehicle's current position and the intersection equals the second distance; the third warning timing is when the straight-line distance between the vehicle's current position and the intersection equals the third distance; and the fourth warning timing is when the straight-line distance between the vehicle's current position and the intersection equals the fourth distance, with the first, second, third, and fourth distances increasing sequentially. Alternatively, the first warning timing is when the vehicle arrives at the intersection with the first remaining time; the second warning timing is when the vehicle arrives at the intersection with the second remaining time; the third warning timing is when the vehicle arrives at the intersection with the third remaining time; and the fourth warning timing is when the vehicle arrives at the intersection with the fourth remaining time, with the first, second, third, and fourth times increasing sequentially.

[0012] According to another aspect of the present invention, a vehicle is provided, the vehicle comprising: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores a computer program executable by the at least one processor, the computer program being executed by the at least one processor to enable the at least one processor to perform the risk warning method of any embodiment of the present invention.

[0013] According to another aspect of the present invention, a computer-readable storage medium is provided, which stores computer instructions for causing a processor to execute and implement the risk warning method of any embodiment of the present invention.

[0014] According to another aspect of the present invention, a computer program product is provided, the computer program product including a computer program, which, when executed by a processor, implements the risk warning method of any embodiment of the present invention.

[0015] The technical solution of this invention determines several intersections that the vehicle will pass through using navigation information during vehicle travel. For each intersection, it acquires pedestrian flow data, vehicle flow data, and meteorological factors when the vehicle arrives at the intersection. Then, it combines the vehicle's turning operation type at the intersection and historical accident data for that intersection to construct a feature vector. This feature vector is then input into a pre-trained prediction model to obtain the risk parameters of the vehicle upon arrival at the intersection, thereby executing the corresponding warning operation. Firstly, by pre-determining the intersections the vehicle will pass through using navigation information and obtaining the risk parameters of the vehicle upon arrival at each intersection beforehand, and executing the corresponding warning operation, it achieves a shift from "passive response" warning to "active prediction" warning, providing drivers with more reaction time. This is particularly suitable for high-speed driving and complex intersection scenarios, effectively reducing the probability of traffic accidents at intersections and ensuring driving safety. Secondly, the feature vector is constructed based on multi-dimensional information such as traffic flow data, meteorological factors, driving behavior characteristics, and historical accident data at intersections, overcoming the problems of biased judgment and insufficient accuracy caused by warnings based on a single type of data. The combined input of multi-dimensional features enables the prediction model to more comprehensively and realistically reflect the overall risk level of the intersection, reducing the false alarm rate and the missed alarm rate, and improving the reliability of the early warning results. Thirdly, this invention can achieve fully automated operation from data collection, prediction, decision-making to output, without requiring manual settings or calibration by the driver, and has wide applicability.

[0016] It should be understood that the description in this section is not intended to identify key or essential features of the embodiments of the present invention, nor is it intended to limit the scope of the invention. Other features of the invention will become readily apparent from the following description. Attached Figure Description

[0017] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0018] Figure 1 This is a flowchart illustrating a risk warning method provided in Embodiment 1 of the present invention; Figure 2 This is a flowchart illustrating a risk warning method provided in Embodiment 2 of the present invention; Figure 3 This is a schematic diagram of the structure of a risk warning device provided in Embodiment 3 of the present invention; Figure 4This is a schematic diagram of another risk warning device provided in Embodiment 3 of the present invention; Figure 5 This is a structural schematic diagram of a vehicle provided in Embodiment 4 of the present invention. Detailed Implementation

[0019] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0020] It should be noted that the terms "first," "second," "third," "historical," "target," etc., used in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that embodiments of the invention described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0021] Furthermore, the user-related information involved in the following embodiments of the present invention is information and data authorized by the user or fully authorized by all parties. The collection, storage, use, processing, transmission, provision, disclosure, and application of the relevant data all comply with the relevant laws, regulations, and standards of the relevant countries and regions, take necessary confidentiality measures, do not violate public order and good morals, and provide corresponding operation entry points for users to choose to authorize or refuse.

[0022] Example 1 Figure 1 This is a flowchart illustrating a risk warning method according to Embodiment 1 of the present invention. This embodiment is applicable to scenarios requiring risk warning in assisted driving or autonomous driving scenarios. The method can be executed by a risk warning device, which can be implemented in hardware and / or software. This risk warning device can be configured in a vehicle. Figure 1 As shown, the method includes: S110. During the vehicle's journey, based on navigation information, determine several intersections that the vehicle will pass through.

[0023] In this invention, navigation information can be obtained through an in-vehicle navigation system or through a navigation application installed on the driver's or passenger's mobile terminal. Navigation information includes, but is not limited to, the vehicle's current location, the preset destination, the planned navigation route, the sequence of roads traversed by the route, and road topology information. Road topology information includes the connectivity between roads, intersection coordinates, intersection types (such as crossroads, T-junctions, T-shaped intersections, roundabouts, etc.), and the sequential number of each intersection on the planned navigation route.

[0024] Specifically, after obtaining navigation information, the vehicle can analyze the planned route using its current location as a reference point to obtain the route's latitude and longitude sequence. This sequence is then matched with a high-precision map to map the route's latitude and longitude sequence to the nearest intersection node, thereby determining the IDs and coordinates of the intersections along the route. These intersections are arranged sequentially according to the order in which the vehicle arrives.

[0025] In one embodiment, the number of intersections can be a preset value, such as 3 or 5; or it can be dynamically determined based on the current driving speed, the memory status of the risk warning device, etc.

[0026] In one embodiment, navigation information can be updated in real time. Once the navigation information is updated, the vehicle needs to re-determine several intersections it will pass through to ensure that these intersections always align with the vehicle's actual travel path.

[0027] S120. For each intersection, obtain pedestrian flow data, vehicle flow data, and weather factors when vehicles arrive at the intersection.

[0028] In this invention, pedestrian flow data is used to characterize the number or density of pedestrians, non-motorized vehicles, or other traffic participants within a preset range (e.g., within 50 meters or 100 meters in each direction of the intersection) around the intersection at the time when vehicles are expected to arrive. Vehicle flow data is used to characterize the traffic flow information of motor vehicles in each direction at the intersection at the time when vehicles are expected to arrive. Meteorological factors are used to characterize the meteorological environmental conditions at the intersection at the time when vehicles are expected to arrive.

[0029] In one embodiment, pedestrian / vehicle traffic data can be predicted based on models such as Spatial-Temporal Graph Convolutional Networks (ST-GCN) / Long Short-Term Memory (LSTM). The input to the model can be traffic statistics / operator signaling data / navigation traffic data collected by roadside nodes (such as fixed high-definition cameras and millimeter-wave radars deployed on road infrastructure such as streetlights and traffic signal poles).

[0030] In one embodiment, the meteorological factor can be determined based on real-time meteorological data or weather forecasts. The value range of the meteorological factor is [0,1].

[0031] It should be noted that pedestrian flow data, vehicle flow data, and meteorological factors are all time-sensitive. When the deviation between the actual arrival time of a vehicle at the intersection and its estimated arrival time exceeds a preset time deviation threshold (such as 1 minute, 3 minutes, or 5 minutes), these data need to be re-acquired to ensure the accuracy of subsequent risk parameters.

[0032] S130. Construct a feature vector based on pedestrian flow data, vehicle flow data, meteorological factors, vehicle turning operation types at the intersection, and historical accident data at the intersection.

[0033] In one embodiment, the feature vector can be a five-dimensional vector, where the five dimensions correspond to pedestrian flow data, vehicle flow data, meteorological factors, vehicle turning operation type at the intersection, and historical accident data at the intersection, respectively.

[0034] Specifically, the method of "constructing a feature vector based on pedestrian flow data, vehicle flow data, meteorological factors, vehicle turning operation type at the intersection, and historical accident data at the intersection" in step S130 can include the following three steps.

[0035] Step 1: Normalize the pedestrian flow data, vehicle flow data, and historical accident data of the intersection respectively.

[0036] Since pedestrian flow data, vehicle flow data, and historical accident data have different dimensions and numerical ranges, in order to ensure the accuracy of subsequent risk parameter predictions, the Min-MaxNormalization method can be used to normalize the pedestrian flow data, vehicle flow data, and historical accident data of the intersection, respectively. After normalization, the value range of the pedestrian flow data, vehicle flow data, and historical accident data of the intersection is all [0,1].

[0037] Step 2: Determine the steering factor based on the type of steering operation performed by the vehicle at the intersection.

[0038] The turning maneuvers of vehicles at this intersection include, but are not limited to, at least one of the following: going straight, turning left, turning right, making a U-turn, keeping to the left, or keeping to the right. Since the turning maneuver type is a discrete categorical variable, it cannot be directly input into the prediction model as a numerical feature. Therefore, it needs to be converted into a corresponding turning factor to achieve the numerical transformation of categorical features.

[0039] For example, a mapping table between steering operation types and steering factors can be pre-established, and the steering factor can be determined based on the mapping table.

[0040] The value range of the steering factor is [0,1].

[0041] Step 3: Construct a feature vector based on the turning factor, meteorological factor, and normalized pedestrian flow data, vehicle flow data, and historical accident data of the intersection.

[0042] Based on the above description, the values ​​of the turning factor, meteorological factor, and normalized pedestrian flow data, vehicle flow data, and historical accident data of the intersection are all within the range of [0,1], which means they are mapped to a unified numerical range. This eliminates the dimensional differences between different features, improves the prediction efficiency of the prediction model, and ensures that each dimension of the features can obtain a balanced contribution in the model.

[0043] S140. Input the feature vector into the pre-trained prediction model to obtain the risk parameters of the vehicle when it arrives at the intersection.

[0044] In this invention, the prediction model can adopt any of the following model structures: ST-GCN, LSTM, Transformer architecture, statistical time series model (e.g., Autoregressive Integrated Moving Average (ARIMA), Seasonal Autoregressive Integrated Moving Average (SARIMA)).

[0045] During the training phase of the prediction model, the model learns the mapping relationship between feature vectors from a large amount of historical sample data and known risk parameters to determine the internal weight parameters. During the prediction phase, the currently constructed feature vectors are input into the prediction model, which then outputs the corresponding risk parameters based on the learned mapping relationship.

[0046] Optionally, the prediction model can be updated at a preset update cycle. The update method can be either full retraining or incremental training. This allows the prediction model to continuously adapt to the dynamic changes in the traffic environment, ensuring the accuracy of risk parameter predictions.

[0047] S150, Execute the early warning operation corresponding to the risk parameter.

[0048] In one embodiment, the correspondence between risk parameters and early warning operations can be preset and stored in the risk early warning device in the form of a mapping table, rule base, or decision tree.

[0049] In one possible implementation, the risk parameter is a risk level, such as a first risk level, a second risk level, and a third risk level, with the risk intensity increasing sequentially. The warning action corresponding to the first risk level can be to remain silent (i.e., not to execute the warning), the warning action corresponding to the second risk level can be to remind the driver to pay attention via voice broadcast, and the warning action corresponding to the third risk level can be to trigger the vehicle's active safety control operation.

[0050] In another possible implementation, the risk parameter is a risk score, and the corresponding early warning action is determined based on the relationship between the risk score and a preset threshold.

[0051] The technical solution of this invention determines several intersections that the vehicle will pass through using navigation information during vehicle travel. For each intersection, it acquires pedestrian flow data, vehicle flow data, and meteorological factors when the vehicle arrives at the intersection. Then, it combines the vehicle's turning operation type at the intersection and historical accident data for that intersection to construct a feature vector. This feature vector is then input into a pre-trained prediction model to obtain the risk parameters of the vehicle upon arrival at the intersection, thereby executing the corresponding warning operation. Firstly, by pre-determining the intersections the vehicle will pass through using navigation information and obtaining the risk parameters of the vehicle upon arrival at each intersection beforehand, and executing the corresponding warning operation, it achieves a shift from "passive response" warning to "active prediction" warning, providing drivers with more reaction time. This is particularly suitable for high-speed driving and complex intersection scenarios, effectively reducing the probability of traffic accidents at intersections and ensuring driving safety. Secondly, the feature vector is constructed based on multi-dimensional information such as traffic flow data, meteorological factors, driving behavior characteristics, and historical accident data at intersections, overcoming the problems of biased judgment and insufficient accuracy caused by warnings based on a single type of data. The combined input of multi-dimensional features enables the prediction model to more comprehensively and realistically reflect the overall risk level of the intersection, reducing the false alarm rate and the missed alarm rate, and improving the reliability of the early warning results. Thirdly, this invention can achieve fully automated operation from data collection, prediction, decision-making to output, without requiring manual settings or calibration by the driver, and has wide applicability.

[0052] Example 2 Figure 2 This is a flowchart illustrating a risk warning method provided in Embodiment 2 of the present invention. Based on Embodiment 1, this embodiment provides a detailed warning process and a method for correcting risk parameters. Figure 2 As shown, the method includes: S201. During the vehicle's journey, based on navigation information, determine the several intersections the vehicle will pass through.

[0053] In this invention, navigation information can be obtained through an in-vehicle navigation system or through a navigation application installed on the driver's or passenger's mobile terminal. Specifically, after obtaining the navigation information, the current location of the vehicle can be used as a reference point to analyze the planned navigation route, obtaining a sequence of latitude and longitude coordinates. This sequence is then matched with a high-precision map to map the route's latitude and longitude coordinates to the nearest intersection node, thereby determining the IDs and coordinates of the intersections along the route. These intersections are arranged sequentially according to the order in which the vehicle arrives.

[0054] Furthermore, the estimated timestamp of the vehicle's arrival at each intersection can be calculated based on the Estimated Time of Arrival (ETA) function of the in-vehicle navigation system / navigation application, and the type of turning operation of the vehicle at each intersection can be extracted according to the turning operations in the route (e.g., left turn, right turn, straight, U-turn, etc.).

[0055] After the above processing, for each intersection the vehicle is about to pass through, the following data can be output: {Intersection ID, Longitude, Latitude, Estimated Arrival Timestamp, Turning Operation Type}.

[0056] S202. For each intersection, obtain pedestrian flow data and vehicle flow data when vehicles arrive at the intersection.

[0057] Specifically, the method for "obtaining pedestrian and vehicle traffic data when vehicles arrive at the intersection" in step S202 may include the following three steps: Step A1: Obtain the historical pedestrian density and historical vehicle density of the target area over the past M first time units, where the target area is the area where the vehicle is currently located, and the target area includes all intersections that the vehicle is about to pass through.

[0058] Historical pedestrian and vehicle density can be extracted from traffic statistics data collected from roadside nodes (such as fixed high-definition cameras and millimeter-wave radars deployed on road infrastructure such as streetlights and traffic signal poles), operator signaling data (such as de-identified mobile device location information), and navigation traffic data.

[0059] The road conditions in the target area can be represented by modeling G=(V,E), where V is a set of elements, each element in the set corresponds to an intersection or road segment, and E is a set of edges. If there is a road that directly connects two elements in V, then there is an edge in E that connects these two elements.

[0060] The first time unit is the prediction accuracy of pedestrian / vehicle density. The first time unit can be set according to actual needs.

[0061] Step A2: Input the historical pedestrian flow density and historical vehicle flow density into the spatiotemporal graph convolutional network to obtain the predicted pedestrian flow density and predicted vehicle flow density of the target area in the next N first time units, where M and N are both positive integers and M>N.

[0062] The reason for setting M > N is that a longer historical time window can provide sufficient contextual information for the spatiotemporal graph convolutional network, enabling the network to fully learn the spatial relationships between intersections and the changing patterns of pedestrian and vehicle density over time, thereby improving the accuracy of predicting pedestrian and vehicle density within the next N first time units.

[0063] Assuming M is 12, N is 6, and the first time unit is 5 minutes, then the input (historical pedestrian density / historical vehicle density) of the spatiotemporal graph convolutional network is the data sampled every 5 minutes in the past hour, and the output (predicted pedestrian density / predicted vehicle density) of the spatiotemporal graph convolutional network is the data sampled every 5 minutes in the next half hour.

[0064] Pedestrian density and vehicle density are correlated over time (e.g., they increase simultaneously during morning and evening rush hours). Therefore, predicted pedestrian density and predicted vehicle density can be output separately or combined into a dual-channel output.

[0065] Step A3: Calculate pedestrian flow data based on the intersection's latitude and longitude information, vehicle arrival time at the intersection, and predicted pedestrian density; and calculate traffic flow data based on the intersection's latitude and longitude information, vehicle arrival time at the intersection, and predicted traffic density.

[0066] After obtaining the predicted pedestrian density and predicted vehicle density of the target area in the next N time units, the predicted pedestrian density and predicted vehicle density of the intersection in the next N time units can be determined based on the latitude and longitude information of the intersection. Then, based on the time when the vehicles arrive at the intersection, the predicted pedestrian density and predicted vehicle density of the two adjacent time units in the next N time units are taken and linearly interpolated to obtain the pedestrian flow data and vehicle flow data.

[0067] S203. For each intersection, obtain the meteorological factors when the vehicle arrives at the intersection.

[0068] In one possible implementation, the method for "obtaining the weather factors when the vehicle arrives at the intersection" in step S203 may include the following three steps: Step B1: Obtain the meteorological forecast data for the intersection in the next P second time units, where P is a positive integer.

[0069] The second time unit represents the forecast accuracy of meteorological factors. This second time unit can be set according to actual needs. Optionally, the first and second time units can be the same or different.

[0070] Specifically, meteorological forecast data should include at least rainfall intensity, visibility, and haze concentration.

[0071] Step B2: Calculate the perception reliability coefficient for each second time unit based on the meteorological forecast data for each second time unit.

[0072] The perceived reliability coefficient is calculated based on the following formula: ;in, The value represents rainfall intensity (in mm / h), and 150 is the maximum impact threshold for rainfall intensity. The weights corresponding to rainfall intensity Visibility is measured in meters (m), and 200 is the maximum visibility threshold. The weight corresponding to visibility. Haze concentration (unit: μg / m³) 3 ), 100 is the maximum impact threshold of smog concentration, The weights corresponding to the smog concentration.

[0073] , , The value can be set according to actual needs (for example, the weight corresponding to rainfall intensity can be increased in areas with significant rainfall impact), and the sum of the three is 1.

[0074] For example, suppose , , The values ​​of α are 0.4, 0.4, and 0.2. When the weather is sunny (rain=0, visibility=1000, fog=0), α = 0.4×(1-0) + 0.4×1 + 0.2×(1-0) = 0.4+0.4+0.2= 1.0; when it is raining heavily (rain=30, visibility=150, fog=0), α = 0.4×(1-0.6) + 0.4×0.75+ 0.2×(1-0) = 0.16+0.3+0.2 = 0.66; when it is foggy (rain=0, visibility=1000, fog=100), α = 0.4×1 + 0.4×1 + 0.2×(1-1) = 0.8.

[0075] Step B3: Calculate the meteorological factors based on the vehicle's arrival time at the intersection and the perception reliability coefficient.

[0076] After obtaining the perception reliability coefficient corresponding to each second time unit, the meteorological factor can be obtained by linearly interpolating the perception reliability coefficients of the two adjacent second time units from the future P second time units based on the time when the vehicle arrives at the intersection.

[0077] Meteorological factors can be expressed as The meteorological factor ranges from [0,1]. A larger value indicates worse weather, while a smaller value indicates better weather.

[0078] In another possible implementation, the perception reliability coefficient in step B2 above can be directly determined based on the discrete assignment of meteorological forecast data: the weather is divided into 7 categories: "sunny, cloudy, light rain, moderate rain, heavy rain, snow, and fog," and each category is assigned a fixed perception reliability coefficient. For example, sunny = 1.0, cloudy = 0.9, light rain = 0.8, moderate rain = 0.6, heavy rain = 0.4, snow = 0.3, and fog = 0.2.

[0079] It should be noted that there is no specific order of execution between steps S202 and S203. That is, step S202 can be executed first and step S203 can be executed later; step S203 can be executed first and step S202 can be executed later; or step S202 and step S203 can be executed simultaneously.

[0080] S204. Normalize the pedestrian flow data, vehicle flow data, and historical accident data of the intersection respectively.

[0081] Since pedestrian flow data, vehicle flow data, and historical accident data have different dimensions and numerical ranges, in order to ensure the accuracy of subsequent risk parameter predictions, the Min-MaxNormalization method can be used to normalize the pedestrian flow data, vehicle flow data, and historical accident data of the intersection, respectively.

[0082] Specifically, the method for normalizing the historical accident data of this intersection can be as follows: Count the number of accidents that occurred at the intersection over a past period, divide this number by the total traffic flow through the intersection, and obtain the corresponding accident rate for that intersection. Then, perform max-min normalization to obtain P_history = (accident_rate - min_rate) / (max_rate - min_rate), with a value range of [0,1]. Here, accident_rate is the accident rate corresponding to this intersection, min_rate is the minimum accident rate among all intersections, and max_rate is the maximum accident rate among all intersections.

[0083] The method for normalizing pedestrian / vehicle traffic data is similar to the method for normalizing historical accident data at this intersection, as described above, and will not be repeated here.

[0084] After normalization, the values ​​of pedestrian flow data, vehicle flow data, and historical accident data at the intersection all fall within the range of [0,1].

[0085] S205. Determine the steering factor based on the type of steering operation performed by the vehicle at the intersection.

[0086] The turning maneuvers of vehicles at this intersection include, but are not limited to, at least one of the following: going straight, turning left, turning right, making a U-turn, keeping to the left, or keeping to the right. Since the turning maneuver type is a discrete categorical variable, it cannot be directly input into the prediction model as a numerical feature. Therefore, it needs to be converted into a corresponding turning factor to achieve the numerical transformation of categorical features.

[0087] For example, a mapping table between steering operation types and steering factors can be pre-established, and the steering factor can be determined based on the mapping table. The value range of the steering factor is [0,1].

[0088] The conversion factor reflects the vehicle's trajectory and blind spot pattern at intersections. For example, a U-turn has the longest trajectory and the most severely limited visibility, so the conversion factor for a U-turn is set to 0.9; a left turn requires crossing the oncoming lane and has a large blind spot, so the conversion factor for a left turn is set to 0.8; a right turn has a relatively small blind spot, so the conversion factor for a right turn is set to 0.4; and going straight has the lowest risk, so the conversion factor for going straight is set to 0.9.

[0089] S206. Based on the turning factor, meteorological factor, and normalized pedestrian flow data, vehicle flow data, and historical accident data of the intersection, a feature vector is constructed.

[0090] The feature vector can be represented as X_i(t) = [P_pedestrian(i,t), P_vehicle(i,t), (1-α(t)), P_route(i), P_history(i)]^T. Here, i represents the i-th intersection the vehicle is about to pass through, t represents the time the vehicle arrives at the intersection, P_pedestrian(i,t) represents the normalized pedestrian traffic data, P_vehicle(i,t) represents the normalized vehicle traffic data, (1-α(t)) represents the weather factor, P_route(i) represents the turning factor, and P_history(i) represents the normalized historical accident data for that intersection.

[0091] Based on the above description, the values ​​of the turning factor, meteorological factor, and normalized pedestrian flow data, vehicle flow data, and historical accident data of the intersection are all within the range of [0,1], which means they are mapped to a unified numerical range. This eliminates the dimensional differences between different features, improves the prediction efficiency of the prediction model, and ensures that each dimension of the features can obtain a balanced contribution in the model.

[0092] S207. Input the feature vector into the pre-trained prediction model to obtain the risk parameters of the vehicle when it arrives at the intersection.

[0093] In this invention, the prediction model can adopt any of the following model structures: ST-GCN, LSTM, Transformer architecture, statistical time series model (e.g., ARIMA, SARIMA).

[0094] During the training phase of the prediction model, the model learns the mapping relationship between feature vectors from a large amount of historical sample data and known risk parameters to determine the internal weight parameters. During the prediction phase, the currently constructed feature vectors are input into the prediction model, which then outputs the corresponding risk parameters based on the learned mapping relationship.

[0095] S208. Determine whether the risk parameter is greater than or equal to the preset threshold. If yes, proceed to S209; otherwise, proceed to S212.

[0096] The preset threshold value can be set according to actual needs. For example, the preset threshold value can be 0.6, 0.65, 0.7, 0.75, 0.8, 0.85, 0.9, etc.

[0097] When the risk parameter is greater than or equal to the preset threshold, it indicates that the vehicle is at a high risk at the intersection. At this time, the linkage with the roadside node is triggered to obtain real-time environmental information (i.e., compensation information) to verify and correct the risk parameter, so as to further improve the accuracy of the warning.

[0098] S209. Send a request message to the roadside node at the intersection.

[0099] Optionally, a request message can be sent via the communication module to one or more roadside units (RSUs) covering the intersection.

[0100] S210. Receive real-time environmental data sent by the roadside node based on the request message, and calculate correction parameters based on the real-time environmental data.

[0101] For example, if a dynamic target is detected in the blind spot and may collide with the vehicle, the correction parameter P_real = 0.9; if no dynamic target is detected in the blind spot, or if a dynamic target is detected in the blind spot but may not collide with the vehicle, the correction parameter P_real = 0.1.

[0102] S211. Adjust the risk parameters based on the correction parameters and the communication delay between the vehicle and the roadside node.

[0103] The communication delay between the vehicle and the roadside node determines the weights of the correction parameters and the risk parameters. Based on the correction parameters, the risk parameters, and their respective weights, the corrected risk parameters can be determined.

[0104] Alternatively, the system can detect the driver's level of concentration and use this information to further adjust risk parameters to compensate for insufficient driver attention.

[0105] S212, Execute the early warning operation corresponding to the risk parameter.

[0106] For example, when the risk parameter is less than the first parameter, the vehicle's driving status is recorded at the first warning time; when the risk parameter is greater than or equal to the first parameter and less than the second parameter, the warning information is broadcast by voice at the second warning time; when the risk parameter is greater than or equal to the second parameter and less than the third parameter, the warning information is broadcast by voice at the third warning time, and the warning location is displayed through the center console or head-up display system; when the risk parameter is greater than or equal to the third parameter, the warning information is broadcast by voice at the fourth warning time, and the warning location is displayed through the center console or head-up display system, and a pre-braking signal is sent to the braking system.

[0107] The first warning timing is the moment when the straight-line distance between the vehicle's current position and the intersection equals the first distance; the second warning timing is the moment when the straight-line distance between the vehicle's current position and the intersection equals the second distance; the third warning timing is the moment when the straight-line distance between the vehicle's current position and the intersection equals the third distance; and the fourth warning timing is the moment when the straight-line distance between the vehicle's current position and the intersection equals the fourth distance, with the first, second, third, and fourth distances increasing sequentially. Alternatively, the first warning timing is the moment when the vehicle arrives at the intersection with the first remaining time; the second warning timing is the moment when the vehicle arrives at the intersection with the second remaining time; the third warning timing is the moment when the vehicle arrives at the intersection with the third remaining time; and the fourth warning timing is the moment when the vehicle arrives at the intersection with the fourth remaining time, with the first, second, third, and fourth times increasing sequentially. As can be seen from the risk parameters and corresponding early warning operation examples in Table 1, the early warning timing is set according to the principle of "the higher the risk, the earlier the early warning, and the stronger the measures". The pre-braking signal refers to the brake pre-pressure (pre-boost) signal sent by the vehicle to the braking system through the drive-by-wire interface, such as establishing brake hydraulic pressure in advance, eliminating brake clearance, putting the braking system in a standby state, and shortening the actual braking response time.

[0108] The reason for using the straight-line distance between the vehicle's current location and the intersection to determine whether a warning has been triggered is that the straight-line distance is always less than or equal to the actual navigation path distance, which is equivalent to triggering a warning in advance, giving the driver more time to react. Moreover, the straight-line distance calculation is simpler, does not depend on route selection or map version, and is more stable and reliable in densely trafficked intersections.

[0109] Optionally, a vibration motor can be integrated into the steering wheel or driver's seat to provide warnings through vibrations of different frequencies. Alternatively, the warning can be activated using the color of the ambient lighting inside the vehicle.

[0110] The technical solution of this invention determines several intersections that the vehicle will pass through using navigation information during vehicle travel. For each intersection, it acquires pedestrian flow data, vehicle flow data, and meteorological factors when the vehicle arrives at the intersection. Then, it combines the vehicle's turning operation type at the intersection and historical accident data for that intersection to construct a feature vector. This feature vector is then input into a pre-trained prediction model to obtain the risk parameters of the vehicle upon arrival at the intersection, thereby executing the corresponding warning operation. Firstly, by pre-determining the intersections the vehicle will pass through using navigation information and obtaining the risk parameters of the vehicle upon arrival at each intersection beforehand, and executing the corresponding warning operation, it achieves a shift from "passive response" warning to "active prediction" warning, providing drivers with more reaction time. This is particularly suitable for high-speed driving and complex intersection scenarios, effectively reducing the probability of traffic accidents at intersections and ensuring driving safety. Secondly, the feature vector is constructed based on multi-dimensional information such as traffic flow data, meteorological factors, driving behavior characteristics, and historical accident data at intersections, overcoming the problems of biased judgment and insufficient accuracy caused by warnings based on a single type of data. The combined input of multi-dimensional features enables the prediction model to more comprehensively and realistically reflect the overall risk level of the intersection, reducing the false alarm rate and the missed alarm rate, and improving the reliability of the early warning results. Thirdly, this invention can achieve fully automated operation from data collection, prediction, decision-making to output, without requiring manual settings or calibration by the driver, and has wide applicability.

[0111] Example 3 Figure 3 This is a schematic diagram of a risk warning device provided in Embodiment 3 of the present invention. Figure 3As shown, the device includes: an intersection determination module 301, a data processing module 302, a vector construction module 303, a risk prediction module 304, and an early warning module 305.

[0112] The intersection determination module 301 is used to determine several intersections that the vehicle will pass through based on navigation information during the vehicle's journey. Data processing module 302 is used to acquire pedestrian flow data, vehicle flow data and meteorological factors when vehicles arrive at each intersection. The vector construction module 303 is used to construct feature vectors based on pedestrian flow data, vehicle flow data, meteorological factors, vehicle turning operation types at the intersection, and historical accident data at the intersection. The risk prediction module 304 is used to input the feature vector into the pre-trained prediction model to obtain the risk parameters of the vehicle when it arrives at the intersection. The early warning module 305 is used to execute early warning operations corresponding to the risk parameters.

[0113] Optionally, the data processing module 302 is specifically used to obtain the historical pedestrian flow density and historical vehicle flow density of the target area over the past M first time units, wherein the target area is the area where the vehicle is currently located, and the target area includes all intersections that the vehicle is about to pass through; input the historical pedestrian flow density and historical vehicle flow density into a spatiotemporal graph convolutional network to obtain the predicted pedestrian flow density and predicted vehicle flow density of the target area over the next N first time units, wherein M and N are both positive integers and M > N; calculate pedestrian flow data based on the latitude and longitude information of the intersection, the time when the vehicle arrives at the intersection, and the predicted pedestrian flow density; and calculate vehicle flow data based on the latitude and longitude information of the intersection, the time when the vehicle arrives at the intersection, and the predicted vehicle flow density.

[0114] Optionally, the data processing module 302 is specifically used to acquire meteorological forecast data for the intersection in the next P second time units, where P is a positive integer; calculate the perception reliability coefficient corresponding to each second time unit based on the meteorological forecast data for each second time unit; and calculate the meteorological factor based on the time when the vehicle arrives at the intersection and the perception reliability coefficient.

[0115] Optionally, meteorological forecast data includes rainfall intensity, visibility, and haze concentration; the perception reliability coefficient is calculated based on the following formula: ;in, Rainfall intensity, The weights corresponding to rainfall intensity For visibility, The weight corresponding to visibility. The concentration of smog, The weights corresponding to the smog concentration.

[0116] Optionally, the vector construction module 303 is specifically used to normalize the pedestrian flow data, vehicle flow data, and historical accident data of the intersection respectively; determine the turning factor according to the turning operation type of the vehicle at the intersection; and construct a feature vector based on the turning factor, the meteorological factor, and the normalized pedestrian flow data, vehicle flow data, and historical accident data of the intersection.

[0117] Optional, combined Figure 3 , Figure 4 This is a schematic diagram of another risk warning device provided in Embodiment 3 of the present invention. Figure 4 As shown, the device also includes a risk correction module 306.

[0118] The risk correction module 306 is used to send a request message to the roadside node of the intersection when the risk parameter is greater than or equal to a preset threshold; receive real-time environmental data sent by the roadside node based on the request message, and calculate correction parameters based on the real-time environmental data; and correct the risk parameter based on the correction parameters and the communication delay between the vehicle and the roadside node.

[0119] Optionally, the warning module 305 is specifically used to record the vehicle's driving status at the first warning time when the risk parameter is less than the first parameter; to announce the warning information via voice at the second warning time when the risk parameter is greater than or equal to the first parameter and less than the second parameter; to announce the warning information via voice at the third warning time when the risk parameter is greater than or equal to the second parameter and less than the third parameter, and to display the warning location through the central control panel or head-up display system; and to announce the warning information via voice at the fourth warning time when the risk parameter is greater than or equal to the third parameter, and to display the warning location through the central control panel or head-up display system, and to send a pre-braking signal to the braking system. The first warning time is the moment when the straight-line distance between the vehicle's current position and the intersection is equal to a first distance. The second warning time is the moment when the straight-line distance between the vehicle's current position and the intersection equals the second distance; the third warning time is the moment when the straight-line distance between the vehicle's current position and the intersection equals the third distance; and the fourth warning time is the moment when the straight-line distance between the vehicle's current position and the intersection equals the fourth distance, with the first, second, third, and fourth distances increasing sequentially. Alternatively, the first warning time is the moment when the vehicle arrives at the intersection with the first remaining time; the second warning time is the moment when the vehicle arrives at the intersection with the second remaining time; the third warning time is the moment when the vehicle arrives at the intersection with the third remaining time; and the fourth warning time is the moment when the vehicle arrives at the intersection with the fourth remaining time, with the first, second, third, and fourth times increasing sequentially.

[0120] The risk warning device provided in the embodiments of the present invention can execute the risk warning method provided in any embodiment of the present invention, and has the corresponding functional modules and beneficial effects of the execution method.

[0121] Example 4 Figure 5 This is a schematic diagram of the structure of a vehicle according to Embodiment 4 of the present invention. The vehicle integrates electronic devices, which are intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic devices can also represent various forms of mobile devices. The components shown herein, their connections and relationships, and their functions are merely illustrative and are not intended to limit the implementation of the invention described and / or claimed herein.

[0122] like Figure 5 As shown, vehicle 10 includes at least one processor 11 and a memory, such as read-only memory (ROM) 12 and random access memory (RAM) 13, communicatively connected to at least one processor 11. The memory stores computer programs executable by at least one processor. Processor 11 can perform various appropriate actions and processes based on the computer program stored in ROM 12 or loaded from storage unit 18 into RAM 13. RAM 13 can also store various programs and data required for the operation of vehicle 10. Processor 11, ROM 12, and RAM 13 are interconnected via bus 14. Input / output (I / O) interface 15 is also connected to bus 14.

[0123] Multiple components in vehicle 10 are connected to I / O interface 15, including: input unit 16, such as keyboard, mouse, etc.; output unit 17, such as various types of displays, speakers, etc.; storage unit 18, such as disk, optical disk, etc.; and communication unit 19, such as network card, modem, wireless transceiver, etc. Communication unit 19 allows vehicle 10 to exchange information / data with other devices through computer networks such as the Internet and / or various telecommunications networks.

[0124] Processor 11 can be a variety of general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of processor 11 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various special-purpose artificial intelligence (AI) computing chips, various processors running machine learning model algorithms, a digital signal processor (DSP), and any suitable processor, controller, microcontroller, etc. Processor 11 performs the various methods and processes described above, such as risk warning methods.

[0125] In some embodiments, the risk warning method may be implemented as a computer program tangibly contained in a computer-readable storage medium, such as storage unit 18. In some embodiments, part or all of the computer program may be loaded and / or installed on vehicle 10 via ROM 12 and / or communication unit 19. When the computer program is loaded into RAM 13 and executed by processor 11, one or more steps of the risk warning method described above may be performed. Alternatively, in other embodiments, processor 11 may be configured to perform the risk warning method by any other suitable means (e.g., by means of firmware).

[0126] Various embodiments of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field-programmable gate arrays (FPGAs), application-specific integrated circuits (ASICs), application-specific standard products (ASSPs), systems-on-a-chip (SoCs), payload-programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various embodiments may include implementations in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which may be a dedicated or general-purpose programmable processor, capable of receiving data and instructions from a storage system, at least one input device, and at least one output device, and transmitting data and instructions to the storage system, the at least one input device, and the at least one output device.

[0127] Computer programs used to implement the methods of the present invention may be written in any combination of one or more programming languages. These computer programs may be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device, such that when executed by the processor, the computer programs cause the functions / operations specified in the flowcharts and / or block diagrams to be performed. The computer programs may be executed entirely on a machine, partially on a machine, or as a standalone software package, partially on a machine and partially on a remote machine, or entirely on a remote machine or server.

[0128] In the context of this invention, a computer-readable storage medium can be a tangible medium that may contain or store a computer program for use by or in conjunction with an instruction execution system, apparatus, or device. A computer-readable storage medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination thereof. Alternatively, a computer-readable storage medium may be a machine-readable signal medium. More specific examples of machine-readable storage media include electrical connections based on one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fibers, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof.

[0129] To provide interaction with the user, the systems and technologies described herein can be implemented in a vehicle having: a display device for displaying information to the user (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor); and a keyboard and pointing device (e.g., a mouse or trackball) through which the user provides input to the vehicle. Other types of devices can also be used to provide interaction with the user; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including sound input, voice input, or tactile input).

[0130] The systems and technologies described herein can be implemented in computing systems that include backend components (e.g., as data servers), or middleware components (e.g., application servers), or frontend components (e.g., user computers with graphical user interfaces or web browsers through which users can interact with implementations of the systems and technologies described herein), or any combination of such backend, middleware, or frontend components. The components of the system can be interconnected via digital data communication of any form or medium (e.g., communication networks). Examples of communication networks include local area networks (LANs), wide area networks (WANs), blockchain networks, and the Internet.

[0131] A computing system can include clients and servers. Clients and servers are generally located far apart and typically interact through communication networks. The client-server relationship is created by computer programs running on the respective computers and having a client-server relationship with each other. The server can be a cloud server, also known as a cloud computing server or cloud host, which is a hosting product within the cloud computing service system to address the shortcomings of traditional physical hosts and VPS services, such as high management difficulty and weak business scalability.

[0132] This invention also provides a computer program product, including a computer program that, when executed by a processor, implements the risk warning method provided in any embodiment of this invention.

[0133] In implementing the computer program product, computer program code for performing the operations of this invention can be written in one or more programming languages ​​or a combination thereof. Programming languages ​​include object-oriented programming languages ​​such as Java, Smalltalk, and C++, as well as conventional procedural programming languages ​​such as C or similar languages. The program code can be executed entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving remote computers, the remote computer can be connected to the user's computer via any type of network—including a local area network (LAN) or a wide area network (WAN)—or can be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0134] It should be understood that the various forms of processes shown above can be used, with steps reordered, added, or deleted. For example, the steps described in this invention can be executed in parallel, sequentially, or in different orders, as long as the desired result of the technical solution of this invention can be achieved, and this is not limited herein.

[0135] The specific embodiments described above do not constitute a limitation on the scope of protection of this invention. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this invention should be included within the scope of protection of this invention.

Claims

1. A risk warning method, characterized in that, The method includes: During the vehicle's journey, several intersections that the vehicle will pass through are determined based on navigation information; For each intersection, obtain the pedestrian flow data, vehicle flow data, and weather factors when the vehicles arrive at the intersection; A feature vector is constructed based on the pedestrian flow data, the vehicle flow data, the meteorological factors, the turning operation type of the vehicle at the intersection, and the historical accident data of the intersection. The feature vector is input into a pre-trained prediction model to obtain the risk parameters of the vehicle when it arrives at the intersection. Execute the warning operation corresponding to the risk parameter.

2. The risk warning method according to claim 1, characterized in that, The acquisition of pedestrian and vehicle traffic data when the vehicle arrives at the intersection includes: Obtain the historical pedestrian density and historical vehicle density of the target area over the past M first time units, wherein the target area is the area where the vehicle is currently located, and the target area includes all intersections that the vehicle is about to pass through; The historical pedestrian density and the historical vehicle density are input into a spatiotemporal graph convolutional network to obtain the predicted pedestrian density and predicted vehicle density of the target area in the next N first time units, where M and N are both positive integers and M > N; The pedestrian flow data is calculated based on the latitude and longitude information of the intersection, the arrival time of the vehicles at the intersection, and the predicted pedestrian flow density; and the traffic flow data is calculated based on the latitude and longitude information of the intersection, the arrival time of the vehicles at the intersection, and the predicted traffic flow density.

3. The risk warning method according to claim 1, characterized in that, The acquisition of the meteorological factors when the vehicle arrives at the intersection includes: Obtain the meteorological forecast data for this intersection in the next P second time units, where P is a positive integer; Calculate the perception reliability coefficient for each second time unit based on the meteorological forecast data for each second time unit. The meteorological factor is calculated based on the time the vehicle arrives at the intersection and the perception reliability coefficient.

4. The risk warning method according to claim 3, characterized in that, The meteorological forecast data includes rainfall intensity, visibility, and haze concentration; the perception reliability coefficient is calculated based on the following formula: ; in, Rainfall intensity, The weights corresponding to rainfall intensity For visibility, The weight corresponding to visibility. The concentration of smog, The weights corresponding to the smog concentration.

5. The risk warning method according to claim 1, characterized in that, The feature vector is constructed based on the pedestrian flow data, the vehicle flow data, the meteorological factors, the turning operation type of vehicles at the intersection, and the historical accident data of the intersection, including: The pedestrian flow data, the vehicle flow data, and the historical accident data of the intersection are all normalized. Determine the steering factor based on the type of steering operation performed by the vehicle at the intersection; The feature vector is constructed based on the steering factor, the meteorological factor, the normalized pedestrian flow data, the vehicle flow data, and the historical accident data of the intersection.

6. The risk warning method according to claim 1, characterized in that, After obtaining the risk parameters of the vehicle when it arrives at the intersection, the method further includes: If the risk parameter is greater than or equal to the preset threshold, a request message is sent to the roadside node of the intersection. Receive real-time environmental data sent by the roadside node based on the request message, and calculate correction parameters based on the real-time environmental data; The risk parameters are corrected based on the correction parameters and the communication delay between the vehicle and the roadside node.

7. The risk warning method according to claim 1, characterized in that, The execution of the early warning operation corresponding to the risk parameter includes: When the risk parameter is less than the first parameter, the driving status of the vehicle is recorded at the first warning time. When the risk parameter is greater than or equal to the first parameter and less than the second parameter, a warning message is broadcast verbally at the second warning time. When the risk parameter is greater than or equal to the second parameter and less than the third parameter, the warning information is broadcast verbally at the third warning time, and the warning location is displayed through the central control console or head-up display system. When the risk parameter is greater than or equal to the third parameter, a warning message is broadcast verbally at the fourth warning time, and the warning location is displayed through the central control console or head-up display system, and a pre-braking signal is sent to the braking system. Wherein, the first warning timing is the moment when the straight-line distance between the vehicle's current position and the intersection is equal to a first distance, the second warning timing is the moment when the straight-line distance between the vehicle's current position and the intersection is equal to a second distance, the third warning timing is the moment when the straight-line distance between the vehicle's current position and the intersection is equal to a third distance, and the fourth warning timing is the moment when the straight-line distance between the vehicle's current position and the intersection is equal to a fourth distance, with the first distance, the second distance, the third distance, and the fourth distance increasing sequentially; Alternatively, the first warning time is the time when the vehicle arrives at the intersection with the first remaining time, the second warning time is the time when the vehicle arrives at the intersection with the second remaining time, the third warning time is the time when the vehicle arrives at the intersection with the third remaining time, and the fourth warning time is the time when the vehicle arrives at the intersection with the fourth remaining time, with the first time, the second time, the third time, and the fourth time increasing sequentially.

8. A vehicle, characterized in that, The vehicles include: At least one processor; and a memory communicatively connected to said at least one processor; wherein, The memory stores a computer program that can be executed by the at least one processor, the computer program being executed by the at least one processor to enable the at least one processor to perform the risk warning method according to any one of claims 1-7.

9. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer instructions that are used to cause a processor to execute the risk warning method according to any one of claims 1-7.

10. A computer program product, characterized in that, The computer program product includes a computer program that, when executed by a processor, implements the risk warning method according to any one of claims 1-7.

Images