Road risk prompt information generation method, equipment and medium

Abstract

The invention discloses a road risk prompt information generation method and device and a medium, and relates to the technical field of data processing. The method comprises the steps of obtaining multi-source monitoring data of a current road, and performing mapping processing on the multi-source monitoring data to obtain a first risk level corresponding to each risk event; determining a comprehensive risk score based on the first risk level corresponding to each risk event, and determining a second risk level of the current road based on the comprehensive risk score; performing feature extraction on the multi-source monitoring data to obtain a first feature, and verifying a second risk level of the current road based on the first feature to obtain a first probability of the second risk level of the current road; and if the first probability is greater than or equal to a probability threshold, generating prompt information corresponding to a second risk level of the current road. Therefore, automatic generation of the road risk information can be realized.

Description

Technical Field

[0001] This application relates to the field of data processing technology, and in particular to a method, device and medium for generating road risk warning information. Background Technology

[0002] With the continuous expansion of road networks and the sustained increase in vehicle ownership, real-time information alerts for road emergencies have become an important research topic in the field of traffic management. Failure to promptly alert vehicles to road emergencies can not only escalate the impact of the event but also potentially lead to serious secondary accidents.

[0003] Road information is mainly disseminated through highway information boards on roads. Traditional road emergency information dissemination is primarily achieved through manual editing and publishing, which is time-consuming and susceptible to human factors. Summary of the Invention

[0004] This application provides a method, device, and medium for generating road risk warning information to solve the following technical problem: how to achieve automated generation of road risk information.

[0005] In a first aspect, embodiments of this application provide a method for generating road risk warning information, the method comprising: Acquire multi-source monitoring data of the current road, and perform mapping processing on the multi-source monitoring data to obtain the first risk level corresponding to each risk event; Based on the first risk level corresponding to each risk event, a comprehensive risk score is determined, and based on the comprehensive risk score, a second risk level for the current road is determined; Feature extraction is performed on the multi-source monitoring data to obtain a first feature, and based on the first feature, the second risk level of the current road is verified to obtain a first probability of the second risk level of the current road; If the first probability is greater than or equal to the probability threshold, a prompt message corresponding to the second risk level of the current road is generated.

[0006] Secondly, embodiments of this application also provide a road risk warning information generation device, the device comprising: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, the instructions being executed by the at least one processor to enable the at least one processor to perform the method as described above.

[0007] Thirdly, embodiments of this application also provide a computer storage medium storing computer-executable instructions, which, when executed, implement the method described in any of the above claims.

[0008] The road risk warning information generation method, device, and medium provided in this application have the following beneficial effects: First, by acquiring and mapping multi-source monitoring data, a comprehensive understanding of road conditions can be achieved based on multi-source information. Rapid mapping yields the first risk level for each risk event, reducing the system's real-time computational burden and improving overall response speed. Next, based on the first risk level for each risk event, a comprehensive risk score is determined, and the second risk level for the current road is determined based on this score. This allows for a comprehensive assessment of the overall road risk based on the impact of multiple individual risk events, avoiding the bias of single-factor judgments. Then, feature extraction is performed on the multi-source monitoring data to obtain the first feature. Based on this first feature, the second risk level for the current road is verified, yielding the first probability of the second risk level. This feature extraction and verification process further validates the second risk level, improving the accuracy and reliability of risk assessment and reducing misjudgments. Finally, if the first probability is greater than or equal to a probability threshold, a warning message corresponding to the second risk level for the current road is generated. This ensures that the risk assessment reaches sufficient confidence, automatically generating the corresponding risk level warning message, guaranteeing the accuracy of early warnings and avoiding unnecessary warnings. Attached Figure Description

[0009] The accompanying drawings, which are included to provide a further understanding of this application and form part of this application, illustrate exemplary embodiments and are used to explain this application, but do not constitute an undue limitation of this application. In the drawings: Figure 1 A flowchart illustrating a method for generating road risk warning information provided in this application embodiment; Figure 2 This is a schematic diagram of the internal structure of a road risk warning information generation device provided in an embodiment of this application. Detailed Implementation

[0010] To make the objectives, technical solutions, and advantages of this application clearer, the technical solutions of this application will be clearly and completely described below in conjunction with specific embodiments and corresponding drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of them. Based on the embodiments in this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0011] The technical solutions proposed in the embodiments of this application will be described in detail below with reference to the accompanying drawings.

[0012] Figure 1This document provides a flowchart of a method for generating road risk warning information, as illustrated in an embodiment of this application. This method can be applied to different types of road risk warning scenarios, such as weather events (e.g., fog, rain, snowfall) and road events (e.g., traffic accidents, congestion, emergency lane occupation, wrong-way driving, pedestrian intrusion). Certain input parameters or intermediate results in the process can be manually adjusted to help improve accuracy.

[0013] This application provides a method for generating road risk warning information. It should be noted that the executing entity in these embodiments can be a server or any terminal device with data processing capabilities. For example, the server can be an independent physical server, a server cluster or distributed system composed of multiple physical servers, or a cloud server providing basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communication, middleware services, domain name services, security services, content delivery networks (CDNs), and big data and artificial intelligence platforms. The terminal device can be a smartphone, tablet, laptop, desktop computer, smart speaker, smartwatch, in-vehicle terminal, etc., but is not limited to these.

[0014] like Figure 1 As shown in the figure, the method for generating road risk warning information provided in this application embodiment specifically includes the following steps: Step 101: Obtain multi-source monitoring data for the current road.

[0015] It should be noted that multi-source monitoring data refers to a diverse set of data reflecting real-time road environment and traffic conditions, collected through different types of equipment and systems. Multi-source monitoring data can include meteorological environmental monitoring data and traffic condition monitoring data. Meteorological environmental monitoring data can be collected through road weather station equipment and can include visibility, temperature, rainfall, wind speed, road snow depth, road water depth, and road ice thickness, etc. Traffic condition monitoring data can be collected through roadside traffic sensing equipment and can include vehicle driving data (such as vehicle speed, vehicle distance, etc.), traffic flow data (such as traffic volume, traffic density, etc.), and video data (such as real-time monitoring footage, etc.).

[0016] As an example, suppose we acquire monitoring data for a section of highway (e.g., the G2 highway section from K120+000 to K125+000). Regarding meteorological data, a weather station installed at K122+500 can detect a current visibility of 180 meters, a current temperature of 8℃, and a humidity of 95%. The wind speed sensor at the weather station records a current wind speed of 5 m / s, and the road surface sensor detects a current road surface water depth of 0.2 mm. Regarding traffic data, vehicle speed radars installed at K121+000 and K124+000 can detect an average vehicle speed of 80 km / h. The radar equipment detects an average distance of 60 meters between vehicles. Traffic camera statistics show that the current traffic volume is 70% of the usual level for the same period, with a traffic density of 45 vehicles per kilometer. Furthermore, a high-definition camera at K123+000 can capture road conditions within a range of approximately 200 meters ahead.

[0017] Step 102: Map the multi-source monitoring data to obtain the first risk level corresponding to each risk event.

[0018] It should be noted that the above mapping processing can be achieved through methods such as rule engines, interval hierarchies, weighted scoring, machine learning models, fuzzy logic, and hybrid mapping, without specific limitations here. Risk events refer to event types corresponding to different monitoring data. For example, visibility data can identify fog risk events, wind speed data can identify strong wind risk events, and ground snow depth data can identify ice and snow risk events. Different risk events can be classified into different risk levels based on the size of the monitoring data. For example, visibility less than 50 meters can be identified as Level I fog risk event, visibility greater than 50 meters and less than 100 meters can be identified as Level II fog risk event, visibility greater than 100 meters and less than 200 meters can be identified as Level III fog risk event, and visibility greater than 200 meters can be identified as Level IV fog risk event.

[0019] Here, the risk level of a risk event can be either a higher number (e.g., a level IV risk level is higher than a level II risk level) or a lower number (e.g., a level II risk level is higher than a level IV risk level), without any specific limitation.

[0020] In some embodiments, step 102 described above can be implemented as follows: classifying the multi-source monitoring data to obtain monitoring data corresponding to each risk event; and for each risk event, mapping the monitoring data corresponding to the risk event based on a preset rule table to obtain the first risk level corresponding to the risk event. Thus, classification processing can associate multi-source monitoring data with different risk events, enabling accurate assessment of the risk level of different risk events based on specific data. Simultaneously, mapping the monitoring data according to the preset rule table ensures the consistency and objectivity of risk assessment.

[0021] As an example, suppose the collected meteorological environmental monitoring data are: visibility 180 meters, temperature 5℃, rainfall intensity 1.5 mm / min, wind speed 8 m / s, and road surface water depth 2 mm; and the traffic condition monitoring data are: average vehicle speed 85 km / h, vehicle distance 120 meters, traffic flow 800 vehicles / hour, and traffic density 35 vehicles / km. Then, by classifying and processing the collected multi-source monitoring data, we can obtain the monitoring data corresponding to the fog risk event as: visibility 180 meters, average vehicle speed 85 km / h, and vehicle distance 120 meters; and the monitoring data corresponding to the rainfall risk event as... The data included a rainfall intensity of 1.5 mm / min, visibility of 180 meters, road surface water depth of 2 mm, and traffic flow of 800 vehicles / hour. Subsequently, the pre-defined rule table for fog risk events was consulted: {Level I: visibility less than 50 meters, Level II: visibility greater than or equal to 50 meters and less than 100 meters, Level III: visibility greater than or equal to 100 meters and less than 200 meters, Level IV: visibility greater than 200 meters}. The current monitoring data showed a visibility of 180 meters. According to the pre-defined rule table, this met the criteria for a Level III fog risk event. This was further supported by an average vehicle speed of 85 km / h. With h and a vehicle distance of 120 meters, it can be confirmed that the event meets the criteria for a Level III heavy fog risk event. Next, refer to the preset rule table for rainfall risk events: {Level I: Hourly rainfall intensity of 30.0 mm / h-49.9 mm / h, or minutely rainfall intensity of 2.1 mm / min-3.0 mm / min with visibility reduced to 100m-150m; Level II: Hourly rainfall intensity of 15.0 mm / h-29.9 mm / h, or minutely rainfall intensity of 1.3 mm / min-2.0 mm / min with visibility reduced to 150m-300m; I...} Level II: Rainfall intensity of 10.0 mm / h-14.9 mm / h per hour, or rainfall intensity of 0.8 mm / min-1.2 mm / min per minute with visibility reduced to 300m-500m. The current monitoring data shows a rainfall intensity of 1.5 mm / min and a visibility of 180 meters. According to the preset rule table mapping, the event meets the criteria for a Level II rainfall risk event. Combined with the road surface water depth and traffic flow, the event is confirmed to meet the criteria for a Level II precipitation risk event. Finally, the levels of fog risk event (Level III) and rainfall risk event (Level II) can be obtained.

[0022] In another example, in a road congestion monitoring scenario, assuming the collected traffic condition monitoring data is an average vehicle speed of 50 km / h and a vehicle distance of 25 m; then, the preset rule table for congestion risk events is queried: {Level I: vehicle speed less than 10 km / h, vehicle distance less than 10 m, congestion length greater than 3 km; Level II: vehicle speed greater than 10 km / h and less than 30 km / h, vehicle distance greater than 10 m and less than 20 m, congestion length greater than 1 km and less than 3 km; Level III: vehicle speed greater than 30 km / h and less than 60 km / h, vehicle distance greater than 20 m and less than 40 m}. According to the preset rule table mapping, the criteria for Level III congestion risk events can be obtained.

[0023] Step 103: Determine the comprehensive risk score based on the first risk level corresponding to each risk event.

[0024] It should be noted that the comprehensive risk score can be determined by a weighted summation method, such as assigning different weights to different types of risk events, multiplying the preset scores corresponding to the first risk level of each risk event by their respective weights, and then summing the results to obtain the comprehensive risk score; the maximum value method can also be used to determine the comprehensive risk score, such as taking the maximum value of the preset scores corresponding to the first risk level of all risk events as the comprehensive risk score; the average value method can also be used to determine the comprehensive risk score, such as calculating the average of the preset scores of the first risk level of all risk events as the comprehensive risk score; the specific method can be determined according to actual needs and is not limited here.

[0025] In some embodiments, step 103 described above can be implemented as follows: if there is only one risk event, the risk score of the first risk level corresponding to the risk event is used as the comprehensive risk score; if there are multiple risk events, the risk scores of the first risk level corresponding to each risk event are merged based on the preset weights corresponding to each risk event to obtain a comprehensive risk score. Thus, when there is only one risk event, complex fusion calculations are unnecessary, and a single risk score is directly used, reducing system complexity and improving response speed; when there are multiple risk events, the cumulative effect of multiple risk events can be comprehensively considered, avoiding the limitations of a single perspective and providing a more comprehensive risk assessment.

[0026] It should be noted that the preset weights for each risk event can be obtained by regression analysis of historical data, for example, by analyzing the impact of each risk event on the probability of an accident through multi-source logistic regression.

[0027] As an example, assuming the collected meteorological and environmental monitoring data are visibility of 150 meters, temperature of 5℃, wind speed of 10m / s, no rainfall, and no snow accumulation, the current road includes a level III fog risk event. The risk score corresponding to the level III fog risk event is 2 points. The risk score of 2 points for the first risk level is directly used as the comprehensive risk score, that is, the comprehensive risk score is 2 points.

[0028] In another example, assuming the collected traffic condition monitoring data is an average vehicle speed of 50 km / h and a vehicle distance of 25 m, the current road includes a congestion risk event of level III, and the risk score corresponding to the level III congestion risk event is 4 points. Then, the risk score of 4 points for the first risk level is directly used as the comprehensive risk score, that is, the comprehensive risk score is 4 points.

[0029] In another example, assuming the collected meteorological and environmental monitoring data are visibility of 120 meters, temperature of 15℃, wind speed of 15 m / s, rainfall intensity of 1.8 mm / min, and no snow accumulation, the current road can be mapped to include a level III fog risk event and a level II rainfall risk event, with corresponding risk scores of 2 and 3 points respectively. The preset weight for the fog risk event is 0.65, and the preset weight for the rainfall risk event is 0.35. Based on the preset weights, the risk score of 2 points for the level III fog risk event and the risk score of 3 points for the level II rainfall risk event are weighted and summed to obtain a final comprehensive risk score of 2.65 points.

[0030] In another example, assuming that the current road includes congestion risk event level III and parking risk event level I based on the collected traffic condition monitoring data, with corresponding risk scores of 4 and 3 points respectively, and the preset weight for congestion risk event is 0.7 and the preset weight for parking risk event is 0.3, the risk score of 4 points for congestion risk event level III and the risk score of 3 points for parking risk event level I are weighted and summed according to the preset weights, and the final comprehensive risk score is 3.7 points.

[0031] Step 104: Based on the comprehensive risk score, determine the second risk level of the current road.

[0032] In some embodiments, step 104 described above can be implemented as follows: based on a preset scoring rule table, the comprehensive risk score is mapped to obtain the second risk level of the current road. Thus, the preset scoring rule table provides a unified risk level assessment standard for all road segments, ensuring the comparability of assessment results across different road segments and at different times. Furthermore, the objective rule table mapping reduces subjective judgment bias inherent in manual assessments, making risk assessments more objective and fair. In addition, the same comprehensive risk score will be mapped to the same risk level, guaranteeing the consistency and repeatability of the assessment results.

[0033] As an example, assume the collected meteorological monitoring data are: visibility 120 meters, temperature 5℃, rainfall intensity 1.2 mm / min, wind speed 8 m / s; traffic monitoring data are: average vehicle speed 85 km / h, vehicle distance 100 meters, traffic flow 600 vehicles / hour; and video data is confirmed by cameras along the route to be low visibility and slow-moving vehicles. By mapping and processing the monitoring data, it can be determined that the current road has a Level III fog risk event (risk score of 60 points) and a Level II rainfall risk event (risk score of 40 points). The weight of the fog risk event is 0.6, and the weight of the rainfall event is 0.4. By weighted summation, the comprehensive risk score is 52 points. The preset scoring rule table is {0~20 points are Level I, 21~40 points are Level II, 41 points~70 points are Level III, and 71 points~100 points are Level IV}. According to the preset scoring rule table, the second risk level of the current road is Level III.

[0034] Step 105: Extract features from the multi-source monitoring data to obtain the first feature.

[0035] It should be noted that feature extraction can be various, including extracting statistical features, time-series features, frequency domain features, spatial features, multi-source data fusion features, dimensionality reduction features, etc., and no specific limitation is made here.

[0036] As an example, suppose that by extracting features from the monitoring data, the visibility features can be obtained as follows: F 1. Humidity characteristics are F 2. Temperature characteristics are F 3. Vehicle speed characteristics are F 4. Vehicle distance characteristics are F 5. Traffic flow characteristics are F 6. Subsequently, all the extracted features are fused, such as through weighted summation or feature concatenation, to obtain the fused features. F That is, the first characteristic.

[0037] In some embodiments, the verification of the second risk level of the current road can be achieved through a trained classification model. Before performing step 105 above, the following processing can also be performed: feature extraction from historical monitoring data to obtain second features; prediction of the third risk level of the current road corresponding to the historical monitoring data based on the second features to obtain a second probability of the third risk level of the current road; determination of a first loss based on the second probability and the labels of the historical monitoring data, and training of the initialized classification model based on the first loss to obtain a trained classification model. Thus, through feature extraction from historical data, risk verification, loss calculation, and model training, a complete model learning framework can be constructed, which can not only improve the accuracy and adaptability of risk assessment but also achieve optimized resource allocation and system reliability and stability.

[0038] It should be noted that the first loss can be calculated using the cross-entropy loss function, logarithmic loss function, and focus loss function, without being specifically limited here; the label of historical monitoring data refers to the third risk level of the current road corresponding to the historical monitoring data.

[0039] As an example, raw monitoring data of all winter weather events over the past three years are obtained, and features are extracted from each raw monitoring data point to obtain the second feature. F x By predicting the third risk level of the current road based on the second feature, a second probability can be obtained. p Among them, the second probability p It can characterize the second feature F x The accuracy rate of roads corresponding to the third risk level based on historical monitoring data was then used to calculate the second probability. p The first loss is calculated between the model and the corresponding label. If the label of the historical monitoring data is the third risk level, the probability of the label is 1, otherwise it is 0. The first loss can represent the degree of difference between the model prediction and the actual situation. Finally, the parameters of the model are updated by backpropagation based on the first loss, so as to obtain the trained classification model.

[0040] Here's an explanation of backpropagation: The original monitoring data is input into the input layer of the neural network model (classification model), passes through the hidden layer, and finally reaches the output layer to output the result. This is the forward propagation process of the neural network model. Since there is an error between the output result of the neural network model and the actual result, the error between the result and the actual value is calculated and propagated back from the output layer to the hidden layer until it reaches the input layer. During the backpropagation process, the values ​​of the model parameters are adjusted according to the error. The above process is iterated continuously until convergence.

[0041] Taking the loss function of the classification model in this application embodiment as an example, the server determines the first loss based on the loss function, backpropagates the first loss from the output layer in the classification model, and backpropagates the first loss layer by layer. When the first loss reaches each layer, the gradient (that is, the partial derivative of the loss function with respect to the parameters of each layer) is solved in combination with the propagated first loss, and the corresponding gradient value of the parameters of each layer is updated.

[0042] Step 106: Based on the first feature, verify the second risk level of the current road to obtain the first probability of the second risk level of the current road.

[0043] It should be noted that the verification of the second risk level of the current road can be achieved through the trained classification model mentioned above.

[0044] As an example, suppose on a certain section of highway, based on multi-source monitoring data and by calculating a comprehensive risk score, the current road's second risk level is determined to be Level III. Further verification of this second risk level is needed. Firstly, feature extraction is performed on the multi-source monitoring data to obtain the first feature. F The first feature can include meteorological features, traffic features, and video features extracted from multi-source monitoring data; then, the first feature is classified using a classification model. F Based on the prediction, the probability that the current road's second risk level is III is 0.93.

[0045] Step 107: If the first probability is greater than or equal to the probability threshold, generate a prompt message for the second risk level corresponding to the current road.

[0046] Continuing with the example above, assuming the probability threshold is 0.9, and the first probability of 0.93 is greater than the probability threshold of 0.9, then the following prompt message corresponding to the second risk level of the current road is generated: "Road A, risk level III, relatively high risk, vehicle speed limit 40km / h, pay attention to driving safety".

[0047] In some embodiments, the generation of the prompt information corresponding to the second risk level of the current road in step 107 can be achieved as follows: if the duration of the second risk level of the current road is greater than the first duration, determine a prompt template that matches the second risk level of the current road; map the second risk level of the current road to obtain the speed limit information corresponding to the second risk level of the current road; based on the speed limit information corresponding to the second risk level of the current road, update the replaceable elements in the prompt template to obtain the prompt information corresponding to the second risk level of the current road. Thus, associating the risk level with a specific prompt template ensures that the information content matches the risk level, avoiding over-warning or under-warning. Furthermore, the template update mechanism allows for the generation of personalized prompt information for specific road sections and risk situations, thereby improving the relevance and readability of the information. In addition, continuous duration monitoring enables timely issuance of warning information after a period of risk, effectively avoiding momentary fluctuations.

[0048] As an example, suppose the second risk level of road A is Level III and the duration exceeds 3 minutes (the first duration). A matching prompt template is selected from the preset prompt template library, such as "Road {road} Risk Level III, High Risk, Recommended Vehicle Speed ​​{speed} km / h, Exercise Caution". Then, based on the current road's second risk level III, the system obtains a recommended speed of 60 km / h (the speed limit information corresponding to the current road's second risk level) through a speed limit mapping table. Afterwards, based on the obtained speed limit information, the replaceable objects (e.g., {road}, {speed}) in the prompt template are updated, resulting in the prompt information corresponding to the current road's second risk level: "Road A Risk Level III, High Risk, Recommended Vehicle Speed ​​60 km / h, Exercise Caution".

[0049] In some embodiments, the generation of the prompt information corresponding to the second risk level of the current road in step 107 can be achieved as follows: if the duration of the second risk level of the current road is longer than the first duration, a prompt template matching the second risk level of the current road is determined; the second risk level of the current road and the risk events of the current road are mapped to obtain the speed limit information corresponding to the second risk level of the current road; based on the speed limit information corresponding to the second risk level of the current road and the risk events of the current road, the replaceable elements in the prompt template are updated to obtain the prompt information corresponding to the second risk level of the current road. Thus, associating the risk level with a specific prompt template ensures that the information content matches the degree of risk, avoiding over-warning or under-warning. Furthermore, through the template update mechanism, personalized prompt information for specific road sections and risk situations can be generated, thereby improving the relevance and readability of the information. In addition, through continuous duration monitoring, timely issuance of warning information after a period of risk can effectively avoid momentary fluctuations.

[0050] As an example, assuming the second risk level of road A is Level III and its duration exceeds 3 minutes (the first duration), a matching prompt template is selected from the preset prompt template library, such as "Road {road} risk level III, with {weather} weather, high risk, recommended vehicle speed {speed} km / h, please drive safely". Subsequently, based on the second risk level III of road A and the presence of fog and rain, the system obtains a recommended driving speed of 60 km / h (the speed limit information corresponding to the second risk level of the current road) through the speed limit mapping table. Then, based on the obtained speed limit information and the fog and rain information of road A, the replaceable objects (such as {road}, {speed}, {weather}) in the prompt template are updated, resulting in the prompt information corresponding to the second risk level of the current road as "Road A risk level III, with fog and rain, high risk, recommended vehicle speed 60 km / h, please drive safely".

[0051] In another example, assuming the second risk level of road A is Level II and the duration exceeds 3 minutes (the first duration), a matching prompt template is selected from the preset prompt template library, such as "Road {road} risk level II, there is a {question} at {distance} ahead, recommended driving speed {speed}, pay attention to driving safety"; then, based on the second risk level II of road A and the parking risk of road A, the system can obtain a recommended driving speed of 20km / h (the speed limit information corresponding to the second risk level of the current road) through the speed limit mapping table; then, based on the obtained speed limit information and the parking risk of road A, the replaceable objects in the prompt template (such as {road}, {distance}, {question}, {speed}) are updated, and the prompt information corresponding to the second risk level of the current road can be obtained as "Road A risk level II, there is a traffic accident 3km ahead, there is a parking risk, recommended driving speed 20km / h, pay attention to driving safety".

[0052] In some embodiments, if the fourth risk level of the current road at a first moment is lower than the second risk level, and the duration of the fourth risk level is greater than or equal to the second duration, a warning message corresponding to the fourth risk level of the current road is generated; if the fifth risk level of the current road at a second moment is greater than the second risk level, a warning message corresponding to the fifth risk level of the current road is generated. Thus, when the fourth risk level is detected to be lower than the second risk level, the system does not immediately lower the risk level warning. By setting a duration threshold (the second duration) as a buffer period, instantaneous fluctuations can be avoided. When the fifth risk level is detected to be greater than the second risk level, the system immediately generates a warning message corresponding to the higher risk, without setting a duration requirement, to ensure a rapid response to risk escalation. This ensures that drivers are promptly alerted when risks intensify, thereby improving road safety.

[0053] As an example, suppose that road A detects a fourth risk level of II at the first moment, which is lower than the second risk level of III, and the duration of the fourth risk level (II) has exceeded the second duration (10 minutes). Then the system generates a corresponding prompt message based on the fourth risk level (II). For example, the generated new prompt message could be "The risk level of road A has been reduced from III to II. There is heavy fog and rain. The risk has decreased. The recommended speed for vehicles is 80 km / h. Please pay attention to driving safety."

[0054] In another example, suppose that at the second moment, the current road A detects that the fifth risk level is IV, which is higher than the second risk level III. The system immediately responds by generating the corresponding prompt information based on the fifth risk level (IV). For example, the generated new prompt information could be: "The risk level of road A has been upgraded from III to IV. There is heavy fog and rain. The risk is high. It is recommended to leave the current road as soon as possible to avoid danger. The recommended driving speed is 20km / h. Please pay attention to driving safety."

[0055] In some embodiments, if the first probability is less than a probability threshold, the second risk level of the current road is downgraded. Thus, by using a probability threshold to test the confidence level of the risk assessment results and downgrading risk levels below the probability threshold, misjudgments caused by data fluctuations or outliers can be avoided, effectively reducing excessive warnings due to uncertain data and improving the accuracy of warning information.

[0056] As an example, suppose on a certain section of highway, based on multi-source monitoring data and by calculating a comprehensive risk score, the current road's second risk level is determined to be Level III. Further verification of this second risk level is needed. Firstly, feature extraction is performed on the multi-source monitoring data to obtain the first feature. F The first feature can include meteorological features, traffic features, and video features extracted from multi-source monitoring data; then, the first feature is classified using a classification model. F The prediction yields a probability of 0.75 for the current road's second risk level to be Level III, which is less than the probability threshold of 0.9. Therefore, the second risk level is downgraded to Level II. Consequently, a warning message corresponding to the current road's risk level is generated: "Road A, Risk Level II, High Risk, Vehicle Speed ​​Limit 60km / h, Please Drive Safely."

[0057] In some embodiments, taking a weather event scenario as an example, the multi-source meteorological environmental monitoring data collected on road A shows a visibility of 120 meters, a temperature of 15°C, a wind speed of 15 m / s, a rainfall intensity of 1.8 mm / min, and no snow accumulation. Through mapping, it can be determined that the current road includes a Level III fog risk event (risk score of 60 points) and a Level II rainfall risk event (risk score of 40 points). The weight of the fog risk event is 0.6, and the weight of the rainfall event is 0.4. Through weighted summation, a comprehensive risk score of 52 points is obtained. According to a preset scoring rule table, the second risk level of the current road is determined to be Level III. Subsequently, feature extraction is performed on the multi-source meteorological environmental monitoring data to obtain the first feature. F According to the first feature FVerifying the current road's second risk level, Level III, yields a probability of 0.93, which is greater than the probability threshold of 0.9. Therefore, the corresponding risk level warning message for the current road is generated: "Road A, Risk Level III, with heavy fog and rain, posing a high risk. Recommended speed: 60 km / h. Please drive carefully."

[0058] In some embodiments, taking a road event scenario as an example, the traffic condition monitoring data collected on road A shows an average vehicle speed of 50 km / h, a vehicle distance of 25 m, and a traffic accident occurring 3 km from the current location. Through mapping, it can be determined that the current road includes a Level III congestion risk event (risk score of 40 points) and a Level II traffic accident risk event (risk score of 60 points). The weight of the congestion risk event is 0.3, and the weight of the traffic accident risk event is 0.7. Through weighted summation, a comprehensive risk score of 54 points is obtained. According to a preset scoring rule table, the second risk level of the current road is determined to be Level III. Subsequently, feature extraction is performed on the multi-source traffic condition monitoring data to obtain the first feature. F According to the first feature F Verifying the current road's second risk level, Level III, yields a probability of 0.95, which is greater than the probability threshold of 0.9. Therefore, the corresponding risk level warning message for the current road is generated: "Road A, Risk Level II, there is a traffic accident 3km ahead, and congestion 1km ahead, posing a risk of stopping. Recommended driving speed: 20km / h. Please drive carefully."

[0059] The above are embodiments of the method proposed in this application. Based on the same inventive concept, embodiments of this application also provide a road risk warning information generation device, the structure of which is as follows: Figure 2 As shown.

[0060] Figure 2 This is a schematic diagram of the internal structure of a road risk warning information generation device provided in an embodiment of this application. Figure 2 As shown, the device includes: At least one processor 201; And a memory 202 that is communicatively connected to at least one processor; The memory 202 stores instructions that can be executed by at least one processor. The instructions are executed by at least one processor 201 to enable at least one processor 201 to perform the steps of the method corresponding to any of the above embodiments.

[0061] Some embodiments of this application provide corresponding to Figure 1 A non-volatile computer storage medium stores computer-executable instructions, which are configured to perform the steps of the method corresponding to any of the above embodiments.

[0062] The various embodiments in this application are described in a progressive manner. Similar or identical parts between embodiments can be referred to mutually. Each embodiment focuses on describing the differences from other embodiments. In particular, the embodiments for IoT devices and media are basically similar to the method embodiments, so the description is relatively simple; relevant parts can be referred to the descriptions of the method embodiments.

[0063] The systems, media, and methods provided in this application are one-to-one correspondences. Therefore, the systems and media also have similar beneficial technical effects as their corresponding methods. Since the beneficial technical effects of the methods have been described in detail above, the beneficial technical effects of the systems and media will not be repeated here.

[0064] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0065] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart... Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0066] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0067] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0068] In a typical configuration, a computing device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.

[0069] Memory may include non-persistent storage in computer-readable media, such as random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash RAM. Memory is an example of computer-readable media.

[0070] Computer-readable media include both permanent and non-permanent, removable and non-removable media that can store information using any method or technology. Information can be computer-readable instructions, data structures, modules of programs, or other data. Examples of computer storage media include, but are not limited to, phase-change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, CD-ROM, digital versatile optical disc (DVD) or other optical storage, magnetic tape, magnetic magnetic disk storage or other magnetic storage devices, or any other non-transferable medium that can be used to store information accessible by a computing device. As defined herein, computer-readable media does not include transient computer-readable media, such as modulated data signals and carrier waves.

[0071] It should also be noted that the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0072] The above description is merely an embodiment of this application and is not intended to limit the scope of this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of the claims of this application.

Claims

1. A method for generating road risk warning information, characterized in that, The method includes: Acquire multi-source monitoring data of the current road and perform mapping processing on the multi-source monitoring data to obtain the first risk level corresponding to each risk event; Based on the first risk level corresponding to each risk event, a comprehensive risk score is determined, and based on the comprehensive risk score, a second risk level for the current road is determined; Feature extraction is performed on the multi-source monitoring data to obtain a first feature, and based on the first feature, the second risk level of the current road is verified to obtain a first probability of the second risk level of the current road; If the first probability is greater than or equal to the probability threshold, a prompt message corresponding to the second risk level of the current road is generated.

2. The method according to claim 1, characterized in that, The mapping process of the multi-source monitoring data to obtain the first risk level corresponding to each risk event includes: The multi-source monitoring data is classified and processed to obtain the monitoring data corresponding to each risk event; For each risk event, based on a preset rule table for the risk event, the monitoring data corresponding to the risk event is mapped to obtain the first risk level corresponding to the risk event.

3. The method according to claim 1, characterized in that, The determination of a comprehensive risk score based on the first risk level corresponding to each risk event includes: If the number of risk events is one, the risk score of the first risk level corresponding to the risk event shall be used as the comprehensive risk score. If there are multiple risk events, the risk scores of the first risk level corresponding to each risk event are merged based on the preset weight corresponding to each risk event to obtain the comprehensive risk score.

4. The method according to claim 1, characterized in that, The determination of the second risk level of the current road based on the comprehensive risk score includes: Based on a preset scoring rule table, the comprehensive risk score is mapped to obtain the second risk level of the current road.

5. The method according to claim 1, characterized in that, The verification of the second risk level of the current road can be achieved through a trained classification model; before extracting features from the multi-source monitoring data to obtain the first feature, the method further includes: Feature extraction is performed on historical monitoring data to obtain the second feature; Based on the second feature, the third risk level of the current road corresponding to the historical monitoring data is predicted to obtain the second probability of the third risk level of the current road; Based on the second probability and the labels of the historical monitoring data, a first loss is determined, and the initial classification model is trained based on the first loss to obtain the trained classification model.

6. The method according to claim 1, characterized in that, The generation of the prompt information corresponding to the second risk level of the current road includes: If the duration of the second risk level of the current road is longer than the first duration, determine a prompt template that matches the second risk level of the current road; Map the second risk level of the current road to obtain the speed limit information corresponding to the second risk level of the current road; Based on the speed limit information corresponding to the second risk level of the current road, the replaceable elements in the prompt template are updated to obtain the prompt information corresponding to the second risk level of the current road.

7. The method according to claim 6, characterized in that, The method further includes: If the fourth risk level of the current road at the first moment is less than the second risk level, and the duration of the fourth risk level is greater than or equal to the second duration, a prompt message corresponding to the fourth risk level of the current road is generated. If the fifth risk level of the current road at the second moment is greater than the second risk level, a prompt message corresponding to the fifth risk level of the current road is generated.

8. The method according to claim 1, characterized in that, The method further includes: If the first probability is less than the probability threshold, the second risk level of the current road is downgraded.

9. A road risk warning information generation device, characterized in that, The device includes: At least one processor; And, a memory communicatively connected to the at least one processor; The memory stores instructions that can be executed by the at least one processor to enable the at least one processor to perform the method as described in any one of claims 1-8.

10. A computer storage medium storing computer-executable instructions, characterized in that, When the computer-executable instructions are executed, they implement the method as described in any one of claims 1-8.

Images