Intelligent teaching method and system for road test training

Abstract

The application provides an intelligent teaching method and system for road test training, relates to the technical field of driving training, and obtains real-time driving information and real-time environment information, inputs the real-time driving information and the real-time environment information into a driving risk degree prediction model, obtains real-time prediction risk degree, judges whether the real-time prediction risk degree is in a predetermined risk degree threshold, obtains real-time driving operation video through a vehicle-mounted camera, analyzes the real-time driving operation video to generate real-time teaching decisions, and dynamically optimizes teaching according to the real-time teaching decisions. The technical problem that the road test training mode in the prior art lacks real-time and accurate data support and targeted teaching optimization is solved. The technical effect that real-time driving information and environment data are combined, the driving behavior of students is comprehensively and real-timely evaluated, targeted teaching optimization is performed, and the driving skills of students are improved is achieved.

Description

Technical Field

[0001] This invention relates to the field of driver training technology, and more specifically to an intelligent teaching method and system for road test training. Background Technology

[0002] Current road test training typically relies on traditional teaching methods, including driving practice under the guidance of an instructor, with the instructor providing verbal guidance and corrections based on the student's performance. However, this approach has several limitations. First, the instructor's teaching experience and subjective judgment significantly impact the effectiveness of instruction, potentially leading to variations in teaching outcomes between different instructors and students. Second, traditional road test training cannot provide real-time and comprehensive monitoring of the student's driving behavior and the surrounding environment, making it difficult to promptly identify and correct potential safety hazards. Furthermore, during road test training, students can only rely on verbal feedback from the instructor or post-test evaluations to understand their shortcomings in driving skills, a method lacking sufficient timeliness and accuracy.

[0003] The existing road test training model lacks real-time, accurate data support and targeted teaching optimization techniques. Summary of the Invention

[0004] This application provides an intelligent teaching method and system for road test training, which addresses the technical problem that existing road test training models lack real-time, accurate data support and targeted teaching optimization.

[0005] In view of the above problems, this application provides an intelligent teaching method and system for road test training.

[0006] The first aspect of this application provides an intelligent teaching method for road test training, the method comprising: acquiring real-time driving information, wherein the real-time driving information refers to vehicle driving information obtained by dynamic driving monitoring of the training vehicle through a motion sensing device mounted on the training vehicle; acquiring real-time environmental information, wherein the real-time environmental information refers to environmental information obtained by dynamic environmental monitoring of the training vehicle through an environmental sensing device mounted on the training vehicle; activating a driving hazard prediction model and inputting the real-time driving information and the real-time environmental information into the driving hazard prediction model to obtain a real-time predicted hazard; determining whether the real-time predicted hazard is within a predetermined hazard threshold; if it is not within the predetermined hazard threshold, linking an onboard camera on the training vehicle and acquiring real-time driving operation video through the onboard camera, wherein the real-time driving operation video refers to a monitoring video of the student's driving operation on the training vehicle; analyzing the real-time driving operation video to generate a real-time teaching decision, and dynamically optimizing the teaching for the student based on the real-time teaching decision.

[0007] A second aspect of this application provides an intelligent teaching system for road test training, the system comprising: a driving information acquisition module for acquiring real-time driving information, wherein the real-time driving information refers to vehicle driving information obtained by dynamic driving monitoring of the training vehicle through a motion sensing device mounted on the training vehicle; an environmental information acquisition module for acquiring real-time environmental information, wherein the real-time environmental information refers to environmental information obtained by dynamic environmental monitoring of the training vehicle through an environmental sensing device mounted on the training vehicle; a prediction model activation module for activating a driving hazard prediction model and inputting the real-time driving information and the real-time environmental information into the driving hazard prediction model to obtain a real-time predicted hazard; a predicted hazard judgment module for judging whether the real-time predicted hazard is within a predetermined hazard threshold; a monitoring video acquisition module for judging whether, if it is not within the predetermined hazard threshold, linking the vehicle-mounted camera on the training vehicle and acquiring real-time driving operation video through the vehicle-mounted camera, wherein the real-time driving operation video refers to the monitoring video of the trainee's driving operation on the training vehicle; and a teaching decision generation module for analyzing the real-time driving operation video to generate real-time teaching decisions and dynamically optimizing the teaching for the trainee based on the real-time teaching decisions.

[0008] One or more technical solutions provided in this application have at least the following technical effects or advantages:

[0009] The method provided in this application acquires real-time driving information and real-time environmental information, activates a driving hazard prediction model, and inputs the real-time driving information and real-time environmental information into the driving hazard prediction model to obtain a real-time predicted hazard. It then determines whether the real-time predicted hazard is within a predetermined hazard threshold, links the onboard camera on the training vehicle, and acquires real-time driving operation video through the onboard camera. This real-time driving operation video refers to the monitoring video of the trainee's driving operations on the training vehicle. The method analyzes the real-time driving operation video to generate real-time teaching decisions, and dynamically optimizes the training for the trainee based on these decisions. This achieves the technical effect of comprehensively and in real-time evaluating trainee driving behavior and performing targeted teaching optimization by combining real-time driving information and environmental data, thereby improving trainee driving skills. Attached Figure Description

[0010] 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.

[0011] Figure 1 A flowchart illustrating the intelligent teaching method for road test training provided in this application;

[0012] Figure 2 A schematic diagram of the intelligent teaching system for road test training provided in this application.

[0013] Explanation of reference numerals in the attached diagram: 11. Driving information acquisition module; 12. Environmental information acquisition module; 13. Prediction model activation module; 4. Prediction hazard assessment module; 15. Monitoring video acquisition module; 16. Teaching decision generation module. Detailed Implementation

[0014] This application provides an intelligent teaching method and system for road test training, addressing the technical problem of existing road test training models lacking real-time, accurate data support and targeted teaching optimization. It achieves the technical effect of combining real-time driving information and environmental data to evaluate student driving behavior in real time and optimize teaching accordingly, thereby improving students' driving skills.

[0015] The technical solutions of the present invention will now be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of them. It should be understood that the present invention is not limited to the exemplary embodiments described herein. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are within the scope of protection of the present invention. It should also be noted that, for ease of description, only the parts related to the present invention are shown in the accompanying drawings, not all of them.

[0016] Example 1, as Figure 1 As shown, this application provides an intelligent teaching method for road test training, the method comprising:

[0017] Step S100: Obtain real-time driving information, which refers to vehicle driving information obtained by dynamically monitoring the driving of the training vehicle through the operation sensing device mounted on the training vehicle.

[0018] Specifically, a series of operational sensors are installed on the road test training vehicle. These sensors can monitor the vehicle's dynamic driving status comprehensively and with high precision. These sensors include wheel speed sensors, acceleration sensors, gyroscope sensors, steering angle sensors, etc., all working together to capture the vehicle's motion and trajectory in real time. Furthermore, based on this real-time monitoring of the vehicle's motion and trajectory, real-time driving information is generated, such as the vehicle's current speed, acceleration, steering angle, lane departure, and braking status, reflecting the driver's operational behavior. By utilizing these sensors to dynamically monitor the vehicle's driving, real-time driving information can be obtained, providing comprehensive and accurate basic data support for intelligent teaching methods. This helps trainees promptly identify and correct driving errors, thereby effectively improving the efficiency and effectiveness of road test training.

[0019] Step S200: Obtain real-time environmental information, which refers to environmental information obtained by dynamically monitoring the training vehicle through environmental sensing devices mounted on the training vehicle.

[0020] Specifically, while acquiring real-time driving information of the training vehicle, environmental sensing devices deployed on the vehicle dynamically monitor and collect data related to the vehicle's driving environment. These environmental sensing devices refer to various sensors and monitoring equipment used to collect real-time information about the environment surrounding the training vehicle, such as temperature and humidity sensors, tire pressure sensors, and light intensity sensors. By monitoring and collecting dynamic environmental information of the training vehicle through these devices, real-time environmental information is generated, including road conditions, weather changes, traffic density, and surrounding obstacles. Obtaining this real-time environmental information provides more comprehensive and accurate data support for road test training, enabling teaching decisions to provide precise guidance to trainees based on real-time data. This helps trainees make correct judgments and decisions in complex and changing driving environments, thereby effectively improving their driving skills and ability to cope with emergencies.

[0021] Step S300: Activate the driving hazard prediction model and input the real-time driving information and the real-time environmental information into the driving hazard prediction model to obtain the real-time predicted hazard.

[0022] Specifically, the driving hazard prediction model is a model built on big data and machine learning algorithms. It is obtained through deep learning and training on historical road test training records and can identify and predict the degree of danger in various driving environments. After collecting real-time driving information and real-time environmental information, the pre-trained driving hazard prediction model is activated, and the real-time driving information and real-time environmental information are used as input data for processing. The driving hazard prediction model, based on pre-trained rules, comprehensively considers the student's driving behavior and changes in the external environment, and outputs a real-time predicted hazard. This real-time predicted hazard is a numerical index reflecting the degree of safety risk under current driving behavior and environmental conditions. The higher the real-time predicted hazard, the greater the risk of the current driving. For example, if the driving hazard prediction model detects that the student is driving too fast and the surrounding traffic is congested, or there are dangerous factors such as water accumulation or ice and snow on the road, it will calculate a high hazard, indicating a high risk in the driving behavior. Conversely, if the student drives more steadily and the external environment is good, the hazard will be relatively low. By inputting real-time driving information and real-time environmental information into the driving hazard prediction model, the current driving hazard level can be quickly and accurately assessed, providing intelligent support for subsequent teaching decisions and safety interventions, thereby improving road safety and teaching effectiveness.

[0023] Step S400: Determine whether the real-time predicted risk level is within a predetermined risk level threshold.

[0024] Step S500: If the vehicle is not at the predetermined danger threshold, the vehicle-mounted camera on the training vehicle is activated, and real-time driving operation video is acquired through the vehicle-mounted camera. The real-time driving operation video refers to the monitoring video of the trainee's driving operation on the training vehicle.

[0025] Specifically, a driving hazard prediction model is used to obtain the real-time predicted hazard level of the current driving process. Based on this, the real-time predicted hazard level is compared with a preset hazard threshold. The preset hazard threshold is a value determined comprehensively based on factors such as safety standards, historical data, and training objectives. It represents the tolerable range of driving behavior under specific driving conditions and is used to distinguish between safe driving states and potentially dangerous driving states. If the real-time predicted hazard level is lower than the preset hazard threshold, the trainee's driving behavior is considered to meet safety standards, requiring no further intervention, and the teaching process continues. If the real-time predicted hazard level is not at the preset hazard threshold, i.e., the real-time predicted hazard level is equal to or higher than the preset hazard threshold, it indicates that the current driving behavior carries a high safety risk and potential danger, requiring intervention and optimization.

[0026] Furthermore, when the real-time predicted hazard level is equal to or higher than a predetermined hazard threshold, the onboard camera in the training vehicle is automatically activated. The onboard camera, installed inside the training vehicle, records the student's driving operations, such as hand and foot movements, steering wheel rotation, and the actions of pressing the accelerator and brake pedals. The onboard camera acquires real-time monitoring video of the student's driving operations in the training vehicle. This real-time driving operation video refers to video recording that monitors the student's driving operations throughout the entire process. It features real-time recording and high precision. Real-time recording means the video is recorded in real-time during driving, ensuring a true reflection of the student's driving behavior. High precision means the onboard camera has high resolution, clearly capturing the student's specific operations, ensuring the reliability and analyzability of the video data. This provides instructors or assessors with comprehensive and intuitive feedback on the student's driving behavior, helping students to promptly identify and correct inappropriate behaviors in a real driving environment, improving the relevance and effectiveness of teaching, and ensuring that students can complete training under safe and efficient conditions.

[0027] Step S600: Analyze the real-time driving operation video to generate real-time teaching decisions, and dynamically optimize the teaching for the trainees based on the real-time teaching decisions.

[0028] Specifically, after obtaining real-time driving operation video, the video is analyzed in detail using video processing algorithms and image recognition technology to identify key driving actions and obtain analysis results. For example, image recognition can detect whether the student has improper speed control when turning, whether there is speeding or sudden braking, or whether they have deviated from their lane. Based on the analysis results, real-time teaching decisions are generated. During the decision generation process, the student's behavior is compared to identify potential errors or deficiencies. For example, if the comparison reveals that the student did not slow down in time when approaching an intersection, a teaching decision will be generated reminding the student to slow down and brake earlier. Dynamic teaching optimization based on real-time teaching decisions allows for adjustments to teaching content and strategies, better helping students improve their driving skills and enhance their safety awareness. This teaching model based on real-time data and dynamic feedback not only improves students' learning efficiency but also addresses potential problems they may encounter during actual road test training, contributing to improved driving skills and ensuring the safety and effectiveness of the training process.

[0029] Further, in step S100, real-time driving information is obtained. This real-time driving information refers to vehicle driving information obtained through dynamic driving monitoring of the training vehicle using operating sensors mounted on the training vehicle. It also includes:

[0030] Step S110, the operating sensing device includes one or more of a wheel speed sensor, an acceleration sensor, a gyroscope sensor, a steering angle sensor, a lateral acceleration sensor, a longitudinal acceleration sensor, a yaw rate sensor, a barometric pressure sensor, a liquid level sensor, a temperature sensor, and an oil pressure sensor.

[0031] Specifically, the motion sensing equipment is used to monitor and collect various motion state information of the vehicle in real time, providing basic data support for the driving hazard prediction model. This motion sensing equipment includes multiple types of sensors, each responsible for collecting different types of vehicle motion data, thereby ensuring a comprehensive and accurate assessment of the vehicle's dynamic state. The motion sensing equipment includes one or more of the following: wheel speed sensors, acceleration sensors, gyroscope sensors, steering angle sensors, lateral acceleration sensors, longitudinal acceleration sensors, yaw rate sensors, air pressure sensors, liquid level sensors, temperature sensors, and oil pressure sensors. The wheel speed sensor measures the rotational speed of each wheel of the vehicle. By monitoring the rotational speed of each tire, the actual driving speed of the vehicle can be obtained, which is used to determine whether the vehicle is speeding or driving smoothly. The acceleration sensor measures the acceleration of the vehicle in different directions. The gyroscope sensor measures the rotational angular velocity of the vehicle, which is used to monitor the rotational changes of the vehicle while driving. The steering angle sensor measures the rotational angle of the vehicle's steering wheel, reflecting the driver's steering operation. The lateral acceleration sensor monitors the lateral force of the vehicle when turning, which can determine whether there is a risk of rollover. The longitudinal acceleration sensor is mainly used to detect changes in longitudinal acceleration when the vehicle accelerates or brakes. The yaw rate sensor measures the rotational rate of the vehicle about its vertical axis, also called yaw angular velocity, which is used to determine whether the vehicle has oversteered or deviated from its driving trajectory during driving. The air pressure sensor is mainly used to detect the air pressure status of the vehicle. The liquid level sensor monitors the changes in the fuel or other liquid levels of the vehicle. The temperature sensor monitors the temperature of key parts of the vehicle in real time, such as engine temperature and brake system temperature. The oil pressure sensor monitors the oil pressure status of the vehicle, especially the oil pressure of the engine and fuel lines. By collecting real-time vehicle operating status data through the onboard sensors, the vehicle's operating condition can be accurately determined, thus providing comprehensive driving data support for the driving hazard prediction model.

[0032] Further, in step S200, real-time environmental information is acquired. This real-time environmental information refers to environmental information obtained through dynamic environmental monitoring of the training vehicle using environmental sensing devices mounted on the training vehicle, and also includes:

[0033] Step S210, the environmental sensing device includes one or more of the following: rain sensor, humidity sensor, temperature sensor, tire pressure sensor, road surface friction coefficient sensor, light intensity sensor, fog light sensor, radar sensor, image sensor, ultrasonic sensor, millimeter wave radar, lidar and air quality sensor.

[0034] Specifically, environmental sensing devices are used to monitor changes in the vehicle's surrounding environment in real time to provide comprehensive environmental information. These devices include one or more of the following: a rain sensor, a humidity sensor, a temperature sensor, a tire pressure sensor, a road surface friction coefficient sensor, a light intensity sensor, a fog light sensor, a radar sensor, an image sensor, an ultrasonic sensor, a millimeter-wave radar, a lidar sensor, and an air quality sensor. The rain sensor detects rainfall around the vehicle, monitors the intensity of rain, and determines whether current weather conditions affect road surface slipperiness, thereby increasing the risk of vehicle skidding. The humidity sensor measures the humidity of the surrounding air, especially in low temperatures, where high humidity can lead to icing or slippage on the road. The temperature sensor monitors the ambient temperature in real time; in cold weather, excessively low temperatures can cause road surface icing, increasing the risk of loss of control. The tire pressure sensor monitors the tire pressure of the vehicle. The road surface friction coefficient sensor detects the current friction force on the road surface; a low friction coefficient indicates the presence of oil, water stains, ice, snow, or other substances on the road surface, potentially causing vehicle skidding. The light intensity sensor monitors the light intensity in the current environment. The system detects whether there is insufficient light; the fog light sensor is mainly used to determine whether the fog lights are in the correct operating state; the radar sensor is used to detect objects or obstacles in front and around the vehicle using electromagnetic waves to compensate for insufficient visibility and identify potential obstacles in a timely manner; the image sensor is used to capture images of the surrounding environment, identify road markings, traffic signs, traffic lights, pedestrians, other vehicles, etc., and assess the complexity of the current traffic environment; the ultrasonic sensor is used to detect obstacles around the vehicle at close range; the millimeter-wave radar and lidar are two high-precision sensing devices that use high-frequency electromagnetic waves and laser beams to scan the surrounding environment, respectively, to accurately identify surrounding obstacles, lane boundaries, and other important landmarks; the air quality sensor monitors the level of air pollution in the surrounding environment. The environmental sensing equipment collects multi-dimensional data of the surrounding environment in real time, including precipitation, humidity, temperature, tire pressure, road friction, light intensity, and air quality, to comprehensively assess the current driving environment, providing timely and accurate driving guidance to trainees and ensuring their safe driving under changing environmental conditions.

[0035] Further, in step S300, before activating the driving hazard prediction model and inputting the real-time driving information and the real-time environmental information into the driving hazard prediction model to obtain the real-time predicted hazard, the method includes:

[0036] Step S301: Obtain historical road test training records based on big data, wherein the historical road test training records include the first historical records;

[0037] Step S302: Compare the first historical driving information in the first historical record with the predetermined driving information to obtain the first comparison result;

[0038] Step S303: The first comparison result after normalization is weighted and calculated to obtain the first comparison deviation index, and the first comparison deviation index is recorded as the first historical risk level.

[0039] Step S304: Extract the first historical environment information from the first historical record, and form a first data group with the first historical driving information and the first historical risk level;

[0040] Step S305: Train the first data set based on the principle of neural network to obtain the driving hazard prediction model.

[0041] Specifically, historical road test training records are obtained from a road test training big data platform. These records include a large amount of past training data from trainees, encompassing their driving and environmental information. A first historical record is selected from these records, containing information on the trainee's driving behavior and environmental conditions under different environments and road conditions during the training process. This first historical driving information is then compared with predetermined driving information, which refers to pre-set driving behaviors representing ideal or standard driving states, such as speed, steering angle, and acceleration under specific road conditions. By comparing the trainee's historical driving information with the predetermined driving information, a first comparison result is obtained, indicating the degree to which the trainee's driving behavior deviates from the predetermined standard. For example, the trainee might drive too fast on a certain section of the road or fail to smoothly control the steering wheel when turning. Next, the first comparison result is normalized to a uniform scale, and a weighted calculation is performed on the normalized first comparison result. Based on historical and expert experience, each comparison result is assigned a different weight according to its impact on driving safety. For example, excessive speed may be more significant than slight steering instability, thus speed-related deviations receive higher weight. A weighted calculation yields a first contrast deviation index, which reflects the degree of deviation between the learner's driving behavior and predetermined driving information, indicating the current level of hazard. After obtaining the first historical hazard level, first historical environmental information is extracted from the first historical data. This first historical environmental information is combined with the first historical driving information and the first historical hazard level to form a complete first data set. This first data set provides comprehensive training data for the neural network, including the learner's driving behavior, external environmental conditions, and their corresponding driving hazard information. Finally, the first data set is trained based on neural network principles. A neural network is a computational model that simulates the working principle of neurons in the human brain, capable of learning and extracting features from large amounts of data to make predictions. During this process, through training, the neural network learns the relationship between different types of driving information, environmental information, and hazard levels, and automatically adjusts the network's weights and parameters, ultimately generating a model capable of predicting driving hazards—the driving hazard prediction model. The driving hazard prediction model can accurately predict the current driving hazard level of the trainee based on new inputs, such as real-time driving information and environmental data, helping trainees to identify potential dangerous behaviors in a timely manner, and providing a scientific basis for subsequent dynamic teaching optimization and safety intervention.

[0042] Further, step S302, before comparing the first historical driving information in the first historical record with the predetermined driving information to obtain the first comparison result, includes:

[0043] Step S302-1: Read the pre-determined road test constraints;

[0044] Step S302-2: Under the predetermined road test constraints, the instructor drives the training vehicle to obtain instructor driving information;

[0045] Step S302-3: Analyze the instructor's driving information to determine the predetermined driving information.

[0046] Specifically, the process involves reading and parsing predetermined road test constraints. These constraints define the driving rules and conditions that trainees must adhere to during road driving skills and safe driving knowledge testing or training. These constraints include speed limits, driving routes, safe distances, steering angles, and acceleration / deceleration requirements. Then, the instructor practices driving the training vehicle under these constraints to ensure the teaching process meets prescribed driving standards and safety requirements. During this process, the instructor follows predetermined routes, speeds, and steering angles, while simultaneously acquiring driving information through sensors mounted on the training vehicle. After the instructor completes the driving and generates driving information, this information is analyzed to extract predetermined driving information that conforms to the predetermined road test constraints, such as standard speeds, steering methods, and braking timing. This predetermined driving information represents the most ideal driving behavior under specific constraints. By analyzing the instructor's driving information under these predetermined constraints, the instructor's driving behavior serves as a benchmark for subsequent teaching, helping trainees understand and follow correct road test driving operations, thus improving the safety and efficiency of road test driving.

[0047] Further, step S600, which involves analyzing the real-time driving operation video to generate real-time teaching decisions and dynamically optimizing teaching for the trainee based on the real-time teaching decisions, also includes:

[0048] Step S610: Extract the first operation video of the first point in the real-time driving operation video, wherein the first point has an identifier of the first associated driving indicator.

[0049] Step S620: Match the first associated index comparison result corresponding to the first associated driving index in the first comparison result;

[0050] Step S630: The first correlation comparison deviation index is obtained by weighting the comparison results of the first correlation index after normalization and recorded as the first correlation risk.

[0051] Step S640: Determine whether the first associated risk level is within the predetermined risk level threshold;

[0052] Step S650: If the driver is not at the predetermined danger threshold, retrieve the instructor's driving operation video and extract the first standard operation video of the first point in the driving operation video.

[0053] Step S660: Use the first standard operation video as the real-time teaching decision.

[0054] Specifically, firstly, the real-time driving operation videos of trainees in training vehicles are analyzed, and specific segments are extracted from the videos, termed the "first operation video at the first point." The first point refers to a specific moment or location where the trainee performs an operation during driving, representing the occurrence of key driving actions such as entering an intersection, turning, or braking. Each point is associated with a specific first-related driving indicator, such as turning speed, stopping distance, or acceleration performance. For example, during the trainee's driving process, a key point is marked, such as the start of braking; the operation at this moment can be identified using the first-related driving indicator, representing the trainee's driving behavior at that moment. Then, the first-related driving indicator at that moment is compared and matched with a first comparison result to obtain the first-related indicator comparison result. This first-related indicator comparison result reflects the degree of difference between the trainee's driving behavior and standard behavior. Next, the first-related indicator comparison results are normalized, transforming all comparison results to a uniform scale. Finally, the normalized first-related indicator comparison results are weighted, assigning a certain weight to each comparison result based on its impact on driving safety. Through weighted calculation, a first correlation comparison deviation index is obtained, which represents the risk level of the trainee at that specific operation point and is denoted as the first correlation risk level. The first correlation risk level reflects the gap between the trainee's driving behavior at that operation point and the safety standard. Then, the first correlation risk level is compared with a predetermined risk threshold to determine whether it falls within that threshold. If it is below the threshold, the trainee's operation meets safety requirements. If the risk level caused by the trainee's driving behavior is equal to or higher than the predetermined threshold, i.e., the first correlation risk level is not within the predetermined risk threshold, it indicates a significant safety risk. The instructor's driving operation video is then retrieved. This video shows the instructor driving the training vehicle under predetermined road test constraints, demonstrating how the instructor operates safely and accurately in the same situation. After retrieving the instructor's driving operation video, the first standard operation video at the same first point is extracted. This standard operation video demonstrates the instructor's standardized operation at that point. Finally, the first standard operation video is used as a real-time teaching decision to provide clear guidance to the trainee. For example, if a student brakes improperly, a video demonstrating the instructor's standard braking technique can be played to show how to brake correctly. By analyzing student driving behavior in detail based on real-time driving operation videos, generating real-time teaching decisions, and dynamically optimizing teaching based on these decisions, precise and personalized teaching feedback can be provided to students. This helps students promptly identify and correct improper driving behaviors, improving driving safety and skill levels.

[0055] Further, in step S640, after determining whether the first associated risk level is at the predetermined risk level threshold, the method further includes:

[0056] Step S641: If the first associated danger level is at the predetermined danger level threshold, extract the second operation video of the second point in the real-time driving operation video, wherein the second point has the identifier of the second associated driving indicator;

[0057] Step S642: Match the second associated index comparison result corresponding to the second associated driving index in the first comparison result;

[0058] Step S643: The comparison results of the second correlation index after normalization are weighted to obtain the second correlation comparison deviation index, which is recorded as the second correlation risk.

[0059] Step S644: Determine whether the second associated risk level is within the predetermined risk level threshold;

[0060] Step S645: If the predetermined danger threshold is not reached, retrieve the instructor's driving operation video and extract the second standard operation video of the second point in the driving operation video;

[0061] Step S646: Use the second standard operation video as the real-time teaching decision.

[0062] Specifically, when the first associated hazard level is within the predetermined hazard threshold, it indicates that the first associated hazard level is within a safe range. The analysis of the student's driving behavior continues, gradually evaluating more operational points. At this stage, the next key point is reached: extracting the second operational video at the second point. The second point represents another critical moment or behavioral node in the student's driving process, such as a turn, sudden braking, or lane change. Each operational point generates a corresponding second associated driving indicator based on the student's behavior and road conditions. Similarly, in the first comparison result, the second associated driving indicator of the student's driving behavior at the second point is matched. By comparing the student's actual driving behavior with the predetermined standard driving behavior, a comparison result of the second associated indicator is generated, indicating whether the student deviated from the standard behavior at the second point. For example, if the student did not appropriately slow down while turning, a deviation of excessive speed will be recorded. The comparison result of the second associated indicator is then normalized to ensure that the dimensions of different indicators are consistent. After normalization, a weighted calculation is performed based on the safety importance of each indicator, i.e., different weights are assigned to highlight the impact of key behaviors. For example, excessive speed while turning may be more dangerous than slight lane departure, therefore speed deviation while turning is given a higher weight. After weighted calculation, a second correlation comparison deviation index is obtained, which reflects the difference between the student's driving behavior and the standard at that moment, and generates a second correlation hazard level. Next, the second correlation hazard level is compared with a predetermined hazard threshold. If the second correlation hazard level is not at the predetermined hazard threshold, it indicates that the student is at a certain risk at that operation point. The instructor's driving operation video is retrieved, and a second standard operation video for the second point is extracted from the driving operation video. The second standard operation video shows the instructor's standard operation in the same situation, providing clear guidance to the student, helping the student understand the correct driving method, and using the second standard operation video as the basis for real-time teaching decisions, providing personalized teaching intervention. Through continuous behavior comparison, hazard calculation, and standard operation demonstration, it is ensured that students can dynamically optimize teaching at each key point, gradually improve driving skills, reduce the risk of accidents, and thus achieve intelligent and refined driving training to improve students' driving skills.

[0063] In summary, the intelligent teaching method for road test training provided in this application embodiment has at least the following technical effects: it combines real-time driving information and environmental data to evaluate the student's driving behavior in real time and optimize the teaching accordingly, thereby improving the student's driving skills and safety.

[0064] Example 2, based on the same inventive concept as the intelligent teaching method for road test training in the foregoing examples, such as... Figure 2 As shown, this application provides an intelligent teaching system for road test training, wherein the system includes:

[0065] The system includes: a driving information acquisition module 11 for acquiring real-time driving information, which refers to vehicle driving information obtained by dynamic driving monitoring of the training vehicle through operation sensing devices mounted on the training vehicle; an environmental information acquisition module 12 for acquiring real-time environmental information, which refers to environmental information obtained by dynamic environmental monitoring of the training vehicle through environmental sensing devices mounted on the training vehicle; a prediction model activation module 13 for activating a driving hazard prediction model and inputting the real-time driving information and the real-time environmental information into the driving hazard prediction model to obtain a real-time predicted hazard; a predicted hazard judgment module 14 for judging whether the real-time predicted hazard is within a predetermined hazard threshold; a monitoring video acquisition module 15 for judging whether it is not within the predetermined hazard threshold, linking the vehicle-mounted camera on the training vehicle, and acquiring real-time driving operation video through the vehicle-mounted camera, wherein the real-time driving operation video refers to the monitoring video of the trainee driving the vehicle on the training vehicle; and a teaching decision generation module 16 for analyzing the real-time driving operation video to generate real-time teaching decisions, and dynamically optimizing the teaching for the trainee based on the real-time teaching decisions.

[0066] Furthermore, the driving information acquisition module 11 includes one or more of the following operating sensing devices: wheel speed sensor, acceleration sensor, gyroscope sensor, steering angle sensor, lateral acceleration sensor, longitudinal acceleration sensor, yaw rate sensor, air pressure sensor, liquid level sensor, temperature sensor, and oil pressure sensor.

[0067] Furthermore, the environmental sensing devices in the environmental information acquisition module 12 include one or more of the following: a rain sensor, a humidity sensor, a temperature sensor, a tire pressure sensor, a road surface friction coefficient sensor, a light intensity sensor, a fog light sensor, a radar sensor, an image sensor, an ultrasonic sensor, a millimeter-wave radar, a lidar, and an air quality sensor.

[0068] Furthermore, the prediction model activation module 13 is also used to perform the following steps: acquiring historical road test training records based on big data, the historical road test training records including a first historical record; comparing the first historical driving information in the first historical record with the predetermined driving information to obtain a first comparison result; performing a weighted calculation on the normalized first comparison result to obtain a first comparison deviation index, and recording the first comparison deviation index as a first historical hazard; extracting the first historical environmental information in the first historical record, and forming a first data group with the first historical driving information and the first historical hazard; training the first data group based on the principle of neural networks to obtain the driving hazard prediction model.

[0069] Furthermore, the prediction model activation module 13 is also used to perform the following steps: reading the predetermined road test constraints; the instructor driving the training vehicle under the predetermined road test constraints to obtain instructor driving information; and analyzing the instructor driving information to determine the predetermined driving information.

[0070] Furthermore, the teaching decision generation module 16 is also used to perform the following steps: extracting a first operation video of a first point in the real-time driving operation video, wherein the first point has an identifier of a first associated driving indicator; matching the first associated indicator comparison result corresponding to the first associated driving indicator in the first comparison result; performing a weighted calculation on the normalized first associated indicator comparison result to obtain a first associated comparison deviation index, and recording it as a first associated hazard; determining whether the first associated hazard is at the predetermined hazard threshold; if it is not at the predetermined hazard threshold, retrieving the instructor's driving operation video, and extracting a first standard operation video of the first point in the driving operation video; and using the first standard operation video as the real-time teaching decision.

[0071] Furthermore, the teaching decision generation module 16 is also used to perform the following steps: if the first associated hazard is at the predetermined hazard threshold, extract the second operation video of the second point in the real-time driving operation video, wherein the second point has an identifier of the second associated driving indicator; match the second associated indicator comparison result corresponding to the second associated driving indicator in the first comparison result; perform a weighted calculation on the normalized second associated indicator comparison result to obtain the second associated comparison deviation index, and record it as the second associated hazard; determine whether the second associated hazard is at the predetermined hazard threshold; if it is not at the predetermined hazard threshold, retrieve the instructor's driving operation video, and extract the second standard operation video of the second point in the driving operation video; use the second standard operation video as the real-time teaching decision.

[0072] The above description is only a preferred embodiment of this application and is not intended to limit this application. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the protection scope of this application.

[0073] This specification and accompanying drawings are merely illustrative examples of this application and are intended to cover any and all modifications, variations, combinations, or equivalents within the scope of this application. Clearly, those skilled in the art can make various alterations and modifications to this application without departing from its scope. Therefore, if such modifications and modifications fall within the scope of this application and its equivalents, this application intends to include such modifications and modifications.

Claims

1. An intelligent teaching method for road test training, characterized in that, include: Acquire real-time driving information, which refers to vehicle driving information obtained by dynamically monitoring the driving of the training vehicle through the operation sensing device mounted on the training vehicle. Acquire real-time environmental information, which refers to environmental information obtained by dynamically monitoring the environment of the training vehicle through environmental sensing devices mounted on the training vehicle. Activating a driving hazard prediction model and inputting the real-time driving information and the real-time environmental information into the model to obtain a real-time predicted hazard includes: acquiring historical road test training records based on big data, the historical road test training records including a first historical record; comparing the first historical driving information in the first historical record with predetermined driving information to obtain a first comparison result; performing a weighted calculation on the normalized first comparison result to obtain a first comparison deviation index, and recording the first comparison deviation index as a first historical hazard; extracting the first historical environmental information from the first historical record, and forming a first data group with the first historical driving information and the first historical hazard; training the first data group based on neural network principles to obtain the driving hazard prediction model. Determine whether the real-time predicted risk level is within a predetermined risk level threshold; If the real-time predicted risk level is equal to or higher than the predetermined risk level threshold, the vehicle-mounted camera on the training vehicle is activated, and real-time driving operation video is obtained through the vehicle-mounted camera. The real-time driving operation video refers to the vehicle operation monitoring video of the trainee on the training vehicle. The real-time driving operation video is analyzed to generate real-time teaching decisions, and the teaching for the trainees is dynamically optimized based on the real-time teaching decisions. This includes: extracting a first operation video of a first point in the real-time driving operation video, wherein the first point has an identifier of a first associated driving indicator; matching the first associated indicator comparison result corresponding to the first associated driving indicator in the first comparison result; performing a weighted calculation on the normalized first associated indicator comparison result to obtain a first associated comparison deviation index, and recording it as a first associated hazard; determining whether the first associated hazard is at a predetermined hazard threshold; if the first associated hazard is equal to or higher than the predetermined hazard threshold, retrieving the instructor's driving operation video and extracting a first standard operation video of the first point in the driving operation video; and using the first standard operation video as the real-time teaching decision.

2. The intelligent teaching method for road test training according to claim 1, characterized in that, The operating sensing devices include one or more of the following: wheel speed sensor, acceleration sensor, gyroscope sensor, steering angle sensor, lateral acceleration sensor, longitudinal acceleration sensor, yaw rate sensor, air pressure sensor, liquid level sensor, temperature sensor, and oil pressure sensor.

3. The intelligent teaching method for road test training according to claim 1, characterized in that, The environmental sensing devices include one or more of the following: rain sensor, humidity sensor, temperature sensor, tire pressure sensor, road friction coefficient sensor, light intensity sensor, fog light sensor, radar sensor, image sensor, ultrasonic sensor, millimeter-wave radar, lidar, and air quality sensor.

4. The intelligent teaching method for road test training according to claim 1, characterized in that, Before comparing the first historical driving information in the first historical record with the predetermined driving information to obtain the first comparison result, the process includes: Read the scheduled road test constraints; Under the constraints of the predetermined road test, the instructor drives the training vehicle and obtains the instructor's driving information; The predetermined driving information is determined by analyzing the instructor's driving information.

5. The intelligent teaching method for road test training according to claim 3, characterized in that, After determining whether the first associated risk level is within the predetermined risk level threshold, the method further includes: If the first associated risk level is lower than the predetermined risk level threshold, extract the second operation video of the second point in the real-time driving operation video, wherein the second point has the identifier of the second associated driving indicator; Match the second associated indicator comparison result corresponding to the second associated driving indicator in the first comparison result; The second correlation comparison deviation index is obtained by weighting the comparison results of the second correlation index after normalization, and is recorded as the second correlation risk. Determine whether the second associated risk level is within the predetermined risk level threshold; If the driver is not at the predetermined danger threshold, retrieve the instructor's driving operation video and extract the second standard operation video of the second point in the driving operation video. The second standard operating procedure video is used as the basis for the real-time teaching decision.

6. An intelligent teaching system for road test training, characterized in that: The intelligent teaching system for road test training is used to execute the steps of any one of the intelligent teaching methods for road test training according to claims 1 to 5, wherein the intelligent teaching system for road test training includes: The driving information acquisition module is used to acquire real-time driving information, which refers to vehicle driving information obtained by dynamically monitoring the driving of the training vehicle through the operation sensing device mounted on the training vehicle. An environmental information acquisition module is used to acquire real-time environmental information, which refers to the environmental information obtained by dynamically monitoring the training vehicle through an environmental sensing device mounted on the training vehicle. The prediction model activation module is used to activate the driving hazard prediction model and input the real-time driving information and the real-time environmental information into the driving hazard prediction model to obtain the real-time predicted hazard. The predicted risk assessment module is used to determine whether the real-time predicted risk is within a predetermined risk threshold; The monitoring video acquisition module is used to determine if the real-time predicted risk level is equal to or higher than the predetermined risk level threshold, and to link the vehicle-mounted camera on the training vehicle to acquire real-time driving operation video through the vehicle-mounted camera. The real-time driving operation video refers to the monitoring video of the trainee's driving operation on the training vehicle. The teaching decision generation module is used to analyze the real-time driving operation video to generate real-time teaching decisions, and to dynamically optimize the teaching for the trainees based on the real-time teaching decisions.

Images