Driving skill training method based on diversified virtual examination room construction

Abstract

The invention discloses a driving skill training method constructed based on a diversified virtual examination room, and relates to the technical field of data processing, and the method comprises the steps: constructing a system based on the diversified virtual examination room, constructing a virtual examination room scene to carry out the driving training of a training user, outputting a driving behavior data set in real time, and carrying out the scoring, and analyzing the data according to the driving score, generating a personalized training path in combination with historical data when the driving score does not reach the standard, and training the user to carry out training step by step according to the training path until a preset skill scoring target is achieved. The technical problems that in the prior art, a traditional driving training method cannot dynamically adjust a training path to meet the actual skill requirements of trainees, consequently, training content lacks pertinence and is low in efficiency are solved, the individuation and pertinence of driving skill training are improved, it is ensured that the training content is highly matched with the actual skill requirements of the trainees, and the driving skill training efficiency is improved. And the training efficiency and effect are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of data processing, and in particular to a driving skill training method based on a diversified virtual examination room. Background Art

[0002] With the rapid development of virtual reality (VR) and artificial intelligence (AI) technologies, traditional driving skills training methods are gradually shifting towards intelligence and personalization. Traditional driving training usually relies on actual vehicles and physical training grounds, which is not only limited by objective conditions such as venues, equipment and weather, but also has the problems of high cost and high risk. In addition, traditional methods are difficult to provide personalized guidance for the specific weaknesses of individual students, and the training effect is often relatively single, unable to fully meet the needs of students with different skill levels, and there are problems of insufficient personalization and lack of dynamic adjustment capabilities. Summary of the invention

[0003] The present application provides a driving skills training method based on a diversified virtual examination room, which is used to solve the technical problem that the traditional driving training method in the prior art cannot dynamically adjust the training path to meet the actual skill needs of the trainees, resulting in a lack of pertinence in the training content and low efficiency.

[0004] The present application provides a driving skill training method based on a diversified virtual examination room construction, the method comprising: obtaining a virtual examination room scene based on a diversified virtual examination room construction system; a training user logs in to the diversified virtual examination room construction system, drives based on the virtual examination room scene, and outputs a real-time driving behavior data set; performs a driving score on the real-time driving behavior data set and outputs a driving score; if the driving score of the real-time driving behavior data set is less than a preset score, obtains a historical driving behavior data set of the training user; inputs the real-time driving behavior data set and the historical driving behavior data set into a driving path generation model, and obtains a first driving training path according to the driving path generation model, wherein the first driving training path includes multiple path nodes, each path node corresponding to a driving training scene; the training user performs driving training according to the first driving training path until the preset score is met.

[0005] One or more technical solutions provided in this application have at least the following technical effects or advantages:

[0006] The driving skill training method based on diversified virtual examination halls provided in the present application relates to the field of data processing technology. The scene is obtained through the virtual examination hall system, and users are allowed to perform driving training, and a driving behavior data set is generated in real time to evaluate the user's performance. When the score does not meet the standard, a personalized training path is generated by combining historical data. The path consists of multiple training scenes. The user performs step-by-step training according to the training path until the preset score target is reached. The technical problem that the traditional driving training method in the prior art cannot dynamically adjust the training path to meet the actual skill needs of the trainees, resulting in a lack of pertinence and low efficiency in the training content is solved. The personalization and pertinence of driving skill training is improved, and the technical effect of ensuring that the training content is highly matched with the actual skill needs of the trainees and improving the efficiency and effectiveness of training is achieved. BRIEF DESCRIPTION OF THE DRAWINGS

[0007] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0008] Figure 1 A flowchart of a driving skills training method based on a diversified virtual examination room provided in an embodiment of the present application;

[0009] Figure 2 A schematic diagram of a process for obtaining a first driving training path in a driving skill training method based on a diversified virtual examination room provided in an embodiment of the present application. DETAILED DESCRIPTION

[0010] The present application provides a driving skills training method based on a diversified virtual examination room, which is used to solve the technical problem that the traditional driving training method in the prior art cannot dynamically adjust the training path to meet the actual skill needs of the trainees, resulting in a lack of pertinence in the training content and low efficiency.

[0011] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of this application.

[0012] It should be noted that the terms "first", "second", etc. in the specification of the present application and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchangeable where appropriate, so that the embodiments of the present application described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, system, product, or server that includes a series of steps or units is not necessarily limited to those steps or units that are clearly listed, but may include other steps or modules that are not clearly listed or inherent to these processes, methods, products, or devices.

[0013] Embodiment 1, as Figure 1 As shown, the present application provides a driving skills training method based on a diversified virtual examination room, the method comprising:

[0014] P10: Build a system based on diversified virtual examination rooms and obtain virtual examination room scenes.

[0015] Among them, the diversified virtual examination room construction system includes multiple construction components, and the multiple construction components include a scene type construction component, a difficulty level construction component, an in-scene task distribution component and an environment simulation component; collaborative construction is performed according to the multiple construction components to obtain a virtual examination room scene.

[0016] Optionally, a virtual examination room scene is obtained based on a diversified virtual examination room construction system. The diversified virtual examination room construction system includes multiple construction components, which work together to complete the generation of virtual examination room scenes. Among them, the scene type construction component selects different types of driving scenes through a preset scene library, such as urban roads, highways, mountain roads, and extreme weather scenes, to ensure the diversity of scene types to meet different driving training needs. The scene type refers to the specific category of the virtual scene, which directly determines the driver's simulation environment, such as traffic signals, road condition complexity, etc.

[0017] The difficulty level building component dynamically adjusts the training intensity of the scene according to the driving skill level, such as increasing traffic flow, introducing dynamic obstacles or simulating complex emergencies. Difficulty level is a standard for grading the training requirements of the virtual test site to adapt to the different needs of beginners and advanced drivers. This component dynamically adjusts parameters based on algorithms to ensure that students can improve their skills in a reasonable challenge.

[0018] The task distribution component in the scenario is used to embed specific driving tasks in the scenario, such as straight-line driving, reversing into a garage, emergency braking, etc., to help students practice targeted skills in the scenario. Task distribution here means allocating pre-set operation goals to the scenario, just like the task points specified in the real driving test.

[0019] The environmental simulation component is responsible for introducing environmental variables, such as lighting conditions, rain and snow, wind speed, and road wetness, and restoring the real driving environment through the support of the physical engine. Environmental simulation is the key to virtual reality technology. Through precise modeling and real-time rendering, drivers can experience a driving experience close to reality.

[0020] Through the coordinated operation of the above-mentioned multiple building components, combined with input parameters and user needs, a virtual test scene that meets the training requirements is finally generated. The data interaction between the components is completed through a unified communication interface to ensure the efficiency and flexibility of the scene construction process. This construction method based on modular design not only improves the richness and accuracy of the virtual test scene, but also provides strong technical support for personalized driving training.

[0021] P20: The training user logs in to the diversified virtual examination room construction system, drives based on the virtual examination room scene, and outputs a real-time driving behavior data set.

[0022] Furthermore, step P20 of the embodiment of the present application further includes:

[0023] P21: The diversified virtual examination room construction system includes a virtual reality operating device, and the training user performs a perceptual sensitivity test according to the virtual reality operating device and outputs a perceptual test data set; P22: When the perceptual test data set meets the preset perceptual sensitivity, the device configuration parameters are obtained; P23: A connection between the virtual reality operating device and the training user is established according to the device configuration parameters.

[0024] It should be understood that the training user logs into the diversified virtual test site construction system, drives based on the virtual test site scene, and outputs a real-time driving behavior data set. At the same time, the user's immersive experience can be enhanced through the virtual reality operation device to ensure the accuracy and consistency of data collection.

[0025] Specifically, before the user practices driving, a perception sensitivity test is first performed. The test content includes the response time and accuracy of multi-modal feedback such as vision, hearing, and touch, and a perception test data set is generated. Perception sensitivity refers to the user's ability to respond to sensory stimuli in a virtual driving environment, such as the visual recognition speed of road signs, the auditory reaction time to emergency alarms, and the sensitivity to tactile feedback of steering wheel vibrations. The test is completed through sensors in virtual reality devices (such as gyroscopes, accelerometers) and external monitoring devices (such as eye trackers).

[0026] After the test is completed, the perception test data set is compared with the preset perception sensitivity. The preset perception sensitivity can be generated based on statistical analysis of user big data. If the data set meets the preset perception sensitivity threshold, the system will obtain the device configuration parameters based on the user's test performance. The device configuration parameters include the display refresh rate, control sensitivity, feedback strength, etc. of the virtual reality device. The dynamic adjustment of these parameters can ensure that the user's operation and feedback in the driving scene are highly synchronized, reducing delays or discomfort.

[0027] Subsequently, the connection between the virtual reality operating device and the training user is established according to the device configuration parameters. This process is completed through the data transmission protocol, ensuring that all operation information in the virtual test room can be synchronized to the user's device in real time and recorded as a real-time driving behavior data set. For example, when the user operates the steering wheel in the virtual scene, the steering angle, force and speed will be accurately recorded and transmitted to the system.

[0028] Through perception sensitivity testing and equipment configuration optimization, the diversified virtual test site construction system ensures that each training user can practice driving in an environment that best suits their personal perception ability, thereby improving the accuracy of data collection and enhancing the training effect. Ultimately, the system records the directional control, acceleration, deceleration, steering and other behaviors generated during driving as a real-time driving behavior data set for subsequent evaluation and analysis.

[0029] P30: Perform a driving score on the real-time driving behavior dataset and output the driving score.

[0030] Specifically, the real-time driving behavior data set is scored and the driving score of the training user is output. The core of this step is to objectively and accurately analyze the driving behavior of the training user to quantify his driving skill level and provide a scientific basis for subsequent training.

[0031] Specifically, the real-time driving behavior data set refers to the user operation data collected during driving, including but not limited to steering wheel angle, braking force, throttle control, speed change, and response time to traffic signals. These data are recorded in real time by high-precision sensors in virtual reality devices and transmitted to the system's evaluation module for analysis.

[0032] First, a variety of key behavior features are extracted from the real-time driving behavior data set, including but not limited to the vehicle's trajectory deviation, acceleration and deceleration smoothness, brake response time, and traffic rules compliance. Here, the trajectory deviation refers to the degree of deviation between the vehicle's driving trajectory and the standard path, which is a key indicator for measuring driving accuracy; acceleration and deceleration smoothness reflects the smoothness and comfort of driving operations, and the data comes from the changes in the force of the accelerator and brake.

[0033] After the feature extraction is completed, the multiple key behavior features are input into the preset driving score model for evaluation. The driving score model can be obtained based on a machine learning algorithm and combined with large-scale driving data training, and can perform adaptability analysis according to the requirements of different driving scenarios and tasks. For example, in urban road scenarios, the model pays more attention to the accuracy of braking and steering, while in highway scenarios, it pays more attention to speed control and lane keeping ability. Exemplarily, the evaluation process can be to first compare the user's real-time driving behavior data with a preset standard driving behavior template, calculate the degree of deviation, and at the same time, assign weights according to the importance of each behavior feature. For example, in complex scenarios, the weight of emergency braking will be higher than other indicators. Then, the weighted sum of each individual score is performed to generate the final driving score.

[0034] The driving score is a quantitative indicator, which can be expressed, for example, from 0 to 100, reflecting the user's overall driving level. The driving score may include a safety score, an operation accuracy score, a rule compliance score, etc. Exemplarily, the safety score can be based on the user's performance in complex traffic scenarios to evaluate their ability to respond to potential dangers. The operation accuracy score can reflect the user's proficiency in the use of control devices such as steering wheels, accelerators, and brakes. The rule compliance score can be scored based on whether the user complies with traffic signals, road markings, and other rules. In addition, a visual evaluation report can be generated to highlight the advantages and disadvantages of the user's operation. For example, if the user's acceleration operation is frequent and not smooth, the instability of the throttle control can be intuitively displayed in the form of a chart, and relevant suggestions can be provided.

[0035] In this way, not only can the driving skills be accurately quantified, but the necessary input data can also be provided for the subsequent customized training path, ensuring the targeted and scientific nature of the driving training process.

[0036] P40: If the driving score of the real-time driving behavior dataset is less than a preset score, obtain the historical driving behavior dataset of the training user.

[0037] Optionally, if the driving score of the real-time driving behavior data set is less than a preset score, the historical driving behavior data set of the training user is obtained. The core of this step is to combine historical driving data with real-time performance to analyze the weak links of the training user's driving skills, thereby providing accurate data support for subsequent training.

[0038] Specifically, the driving score is a quantitative evaluation of the training user's current driving performance, which is usually calculated by weighting multiple evaluation indicators, such as safety score, rule compliance score, operation accuracy score, etc. The preset score is the minimum passing score set by the system, which is based on the driving skills training goals and standards and is used to evaluate whether the user has reached the required driving level.

[0039] When the real-time driving score is lower than the preset score, the system automatically triggers the data retrieval module to obtain the historical driving behavior data set from the user's profile. The historical driving behavior data set is a complete data record collected by the user in previous driving practice, including operating habits, skill improvement trajectories, and past evaluation results. It provides an important basis for identifying the user's driving weaknesses and operating patterns.

[0040] During the data retrieval process, data indexing and classification technology is used to quickly locate historical records related to the current task. For example, when a user performs poorly in a complex turning scenario, the system will prioritize retrieving historical data related to steering control and speed adjustment. By combining comparative analysis of real-time data with historical data, the system can deeply explore the user's operating characteristics, such as repetitive errors, response delays, etc.

[0041] To ensure the comprehensiveness and efficiency of data analysis, historical data is processed based on the driving behavior feature extraction algorithm. Feature extraction refers to extracting key information that can reflect the user's operating habits and skill status from the original data. For example, the system can extract the user's average braking reaction time and steering wheel operation stability indicators in past practice to help locate the source of deviation in current performance.

[0042] By acquiring historical driving behavior data sets, the system can build a complete driving skill profile of the user. This not only provides data support for the design of subsequent training paths, but also identifies the user's skill growth trend, thereby adjusting training goals and strategies, and ultimately achieving targeted and effective driving skill improvement.

[0043] P50: Input the real-time driving behavior dataset and the historical driving behavior dataset into a driving path generation model, and obtain a first driving training path according to the driving path generation model, wherein the first driving training path includes multiple path nodes, and each path node corresponds to a driving training scenario.

[0044] Further, such as Figure 2As shown, step P50 of the embodiment of the present application also includes:

[0045] P51: The driving path generation model extracts driving evaluation item features according to the real-time driving behavior dataset and the historical driving behavior dataset, and obtains evaluation indicators corresponding to each driving evaluation item; P52: Determine the skill items to be improved according to the evaluation indicators corresponding to each driving evaluation item, wherein each driving evaluation item at least includes a braking evaluation item, a speed control evaluation item, a steering operation evaluation item, an environmental adaptability evaluation item, and an emergency response capability evaluation item; P53: Construct a virtual examination room scene library; P54: Perform scene matching in the virtual examination room scene library according to the skill items to be improved, and generate the first driving training path.

[0046] It should be understood that the real-time driving behavior data set and the historical driving behavior data set are input into the driving path generation model, and the first driving training path is obtained according to the driving path generation model. This step generates a personalized training path by combining the user's current driving performance and historical skill data to effectively strengthen the user's weak links. The specific implementation process is as follows.

[0047] First, the driving path generation model performs multi-dimensional analysis and feature extraction based on the input real-time driving behavior data set and historical driving behavior data set. Feature extraction refers to extracting core information that can reflect driving skills and behavior patterns from raw data, such as operating stability, response delay, and rule compliance. In this process, the model will generate corresponding evaluation indicators based on different driving evaluation items (such as braking evaluation items and speed control evaluation items). These indicators quantify the user's specific performance in various skills.

[0048] Based on the above evaluation indicators, by comparing the preset standards and individual performance, the skill items that the user needs to improve, that is, the skill items to be improved, are identified. The skill items to be improved refer to specific abilities that the user performs weakly during driving and require further training, such as delayed braking response or unstable steering operation. The dimensions of the evaluation items cover multiple key skill areas, including but not limited to braking evaluation items, which are used to evaluate whether the timing and strength of the user's braking operation are appropriate. The speed control evaluation item is used to check whether the user's speed adjustment meets the road conditions and task requirements. The steering operation evaluation item is used to evaluate the accuracy and smoothness of the steering operation. The environmental adaptability evaluation item is used to measure the user's performance in different lighting, weather, road conditions and other environments. The emergency response capability evaluation item is used to analyze the user's decision-making speed and operation accuracy in response to emergencies.

[0049] Then, with the support of the virtual test site scene library, targeted training scenes are customized for users. The virtual test site scene library is a scene collection containing a variety of driving tasks, covering a variety of driving environments from urban roads to extreme weather. These scenes are built with highly realistic virtual reality technology and support real-time dynamic loading and scene adjustment.

[0050] Finally, according to the skill items to be improved, scene matching is performed in the virtual test site scene library to generate the first driving training path. Scene matching refers to selecting corresponding driving training scenes from the scene library based on the specific skills that the user needs to improve. For example, for users with unstable steering operations, the system may select complex curve driving scenes as the training focus; for users who react slowly to emergencies, emergency avoidance scenes involving the sudden appearance of obstacles may be designed.

[0051] The first driving training path generated consists of multiple path nodes, each of which corresponds to a specific driving training scenario. These path nodes are arranged in a logical order, gradually guiding users to complete the training goals from strengthening basic skills to improving comprehensive abilities. Through the above steps, the training path can be personalized and targeted, ensuring that each user can efficiently improve driving skills in a scientific training process, laying the foundation for ultimately achieving the preset driving score goals.

[0052] Furthermore, step P53 of the embodiment of the present application also includes:

[0053] P53-1: Set up multiple virtual examination scene samples; P53-2: Collect driving behavior sample data of the training user group based on each virtual examination scene sample; P53-3: Upload the driving behavior sample data of each virtual examination scene sample to the cloud platform, analyze the driving behavior sample data of each virtual examination scene sample on the cloud platform, and determine the skill improvement label of each virtual examination scene sample; P53-4: Label the multiple virtual examination scene samples according to the skill improvement label to generate a virtual examination scene library.

[0054] Optionally, in order to achieve accurate adaptation and scientific configuration of virtual test site scenes, by setting scene samples, collecting user data, analyzing skill improvement effects and scene identification, a systematic virtual test site scene library is finally generated to support the generation of targeted driving training paths. The specific process is as follows:

[0055] First, we create a variety of virtual test scene samples, which cover different driving environments and task requirements, such as urban traffic, mountain roads, highways, rainy and snowy weather, night driving, etc. Virtual test scene samples are simulation test units of different driving environments, used to evaluate students' driving performance under various conditions. Each sample is carefully designed, focusing on the matching of environmental diversity and training objectives.

[0056] After the trainees complete the driving exercises in the above scenario samples, the system collects their driving behavior data in real time. These data include the trainees' operation records in each scenario, such as braking force, acceleration time, steering wheel rotation angle, and response time to emergencies. Through large-scale collection, it can not only reflect individual performance, but also analyze the overall performance trend of the trainee group in a specific scenario.

[0057] All driving behavior sample data is uploaded to the cloud platform and processed using efficient big data analysis technology. The cloud platform uses feature extraction and clustering algorithms to identify the key features of the student's behavior data in each scenario, and analyzes which types of scenario training are most helpful in improving driving safety and which scenarios can effectively improve reaction speed. For example, certain complex road conditions (such as multi-fork intersections) may significantly improve students' ability to comply with rules, while sudden obstacle scenarios have a greater effect on improving reaction time.

[0058] Skill improvement labels are the key results of data analysis. They are quantitative or qualitative descriptions of how each scenario sample improves a trainee's skill. For example, a "driving safety improvement label" may be associated with a complex traffic scenario, and a "reaction speed improvement label" may correspond to an emergency scenario. Through these labels, the system can intuitively understand the training value of the scenario.

[0059] Then, based on the analysis results, a corresponding skill improvement label is attached to each virtual test scene sample, and a standardized scene library is generated. The virtual test scene library is a structured database that stores all analyzed and identified scene samples, and is classified and indexed according to skill improvement goals. For example, a scene for improving reaction speed can be separately identified as an "emergency event intensive training scene", while a scene for improving driving safety may be classified as a "rule compliance intensive scene".

[0060] Through the above steps, the virtual examination scene library has achieved scientific and personalized development, providing rich choices and precise matching support for the generation of subsequent training paths.

[0061] Furthermore, step P54 of the embodiment of the present application also includes:

[0062] P54-1: Generate a model based on the driving path to obtain multiple driving training scenarios; P54-2: Identify the difficulty of the multiple driving training scenarios and output multiple difficulty labels, and the multiple driving training scenarios correspond to the multiple difficulty labels; P54-3: According to the multiple difficulty labels, sequentially connect the multiple driving training scenarios from small to large to generate the first driving training path.

[0063] In a possible embodiment of the present application, in order to achieve the logic and pertinence of the training path, this step is further refined to construct a complete driving training path by generating scenes, difficulty identifiers, and sequential connections. The specific description is as follows:

[0064] First, based on the driving path generation model, combined with the user's real-time driving behavior data and historical driving behavior data, the scenario samples that best match the skill items to be improved are screened from the virtual test site scenario library. These scenarios are designed as specific task units for driving training, such as simulated sharp turn exercises, sudden obstacle avoidance training, or complex traffic rules scenario operations. Driving training scenarios are a collection of specific driving tasks provided to students in the virtual test site, covering multiple skill areas.

[0065] In order to ensure the progressiveness of the training path, the difficulty of the generated driving training scenarios is marked. The difficulty label is a label assigned after a comprehensive evaluation of the operational complexity of the scenario, the dynamic changes of the environment, and the task requirements. For example, a scenario simulating straight-line driving may be marked as "beginner difficulty", while a scenario containing complex curves and emergencies may be marked as "advanced difficulty". The difficulty label is generated by analyzing the historical completion rate of the scenario, the success rate of the trainee operation, and the distribution of key behavior data to ensure its scientificity and rationality.

[0066] According to the difficulty labels, the system sorts multiple driving training scenarios and connects them in order from easy to difficult to form a complete first driving training path. This sorting strategy can help students gradually adapt and master skills. For example, the initial stage of the path may include simple braking and starting practice scenarios, while the subsequent stages gradually transition to complex traffic conditions and advanced operating skills training.

[0067] Each driving training scenario is defined as a path node in the path, and students need to complete the training tasks of each node in sequence. The path nodes are logically connected. For example, after completing a primary-level straight-line driving task, the system will automatically load the next intermediate-level curve driving task. This design ensures the fluency and continuity of training.

[0068] Through the above steps, a scientific and progressive driving training path was successfully constructed. The first driving training path not only covers all the key links of the skills to be improved, but also enhances the confidence of students and improves the learning effect through the design of difficulty gradient. Finally, the sequential connection of the path provides an important technical guarantee for achieving efficient and accurate driving skills improvement.

[0069] P60: The training user performs driving training according to the first driving training path until the preset score is met.

[0070] Specifically, the training user performs driving training according to the first driving training path until the preset score is met. The core of this step is to guide the training user to gradually improve driving skills through a personalized training path, while using real-time feedback and dynamic adjustment mechanisms to ensure that the training effect reaches the expected result.

[0071] At the beginning of training, the system loads the first driving training path, and each node in the path corresponds to a specific driving training scenario. These scenarios are arranged in a progressively difficult form, covering all aspects from basic skills to complex operations. For example, users may start with basic braking and acceleration control exercises, and gradually transition to advanced skill training such as cornering driving and emergency handling.

[0072] In each scenario, the real-time driving behavior acquisition module records the training user's operating behavior, such as steering wheel rotation angle, braking force, throttle control timing, etc. These operating data are transmitted to the driving behavior analysis engine in real time, compared with the preset optimal operating model, and generate instant operation feedback for the trainee. Operation feedback is an evaluation of the user's current behavior, including whether the operation is accurate and whether it complies with driving rules. For example, if the user's directional control is unstable while driving on a curve, the system will immediately issue a prompt and give corrective suggestions through visual or voice methods.

[0073] During the training process, the training intensity is dynamically adjusted according to the user's performance. The dynamic adjustment mechanism is one of the key technologies for training. By analyzing the user's operation data and current status in real time, the system can adjust the training rhythm. For example, when a user fails multiple times in a scene, the system may extend the scene time or reduce the difficulty of the task to ensure that the user gradually masters the skills; conversely, if the user passes a scene quickly, the system will automatically jump to a more difficult training node.

[0074] The user's performance in each scenario is quantified into a scenario score, which is aggregated to generate an overall driving score after the training path is completed. The preset score is the driving skill standard set by the system, which is dynamically adjusted according to the goals of driving training and the user's personal learning progress. For example, a basic student may have a relatively low preset score, while an advanced student may have a higher score requirement.

[0075] When the user's overall driving score reaches or exceeds the preset score, the system determines that the training goal has been achieved. After the training is completed, the system will generate a detailed driving training report, which includes the user's training performance, weak links in operation, and recommended follow-up training suggestions. If the score does not meet the requirements, the system will regenerate an optimized driving training path based on the training results to guide the user to continue to strengthen practice.

[0076] Through this complete training process, users can gradually master driving skills under personalized and scientific guidance, ensuring that they ultimately meet the preset scoring standards, thus laying a solid foundation for actual driving.

[0077] Furthermore, the embodiment of the present application further includes step P70, and step P70 further includes:

[0078] P71: When the training user performs driving training according to the first driving training path, the newly added driving behavior data set of the training user is recorded; P72: the newly added driving behavior data set is input as incremental data into the driving path generation model for feedback learning, the first driving training path is updated, and a second driving path is output; P73: the training user performs driving training according to the second driving training path.

[0079] Optionally, after completing the first driving training path, the embodiment of the present application dynamically optimizes the driving training path through a feedback learning mechanism of incremental data to provide users with a more accurate and effective training experience.

[0080] Specifically, during the process of training the user to perform driving training according to the first driving training path, the system collects the user's new driving behavior data in real time to form a new driving behavior data set. This data set contains all new operation records of the user in the process of completing the training path, such as the accuracy of steering wheel operation, braking reaction time, speed control changes, etc. Compared with historical data, new data can better reflect the user's current skill status and training effect.

[0081] The incremental learning mechanism is used to input new data into the driving path generation model for dynamic optimization. Incremental learning is a machine learning technology that can adjust model parameters and improve the adaptability of the model through new data without retraining the entire model. For example, if the new data shows that the user's operation is unstable in complex traffic scenarios, the model will automatically identify this skill weakness and adjust the subsequent training path.

[0082] During the feedback learning process, the model compares and analyzes the new data with the historical data, identifies the progress and deficiencies of the user's skills, and recalculates the priority and difficulty distribution of the training path in combination with the preset skill improvement goals. Through this optimization method, the model generates a second driving training path, which is not only more targeted, but also can further improve the user's training efficiency.

[0083] The training user then conducts driving training according to the updated second driving training path. Compared with the first path, the second path is closer to the actual needs of users in terms of scenario design and task configuration. For example, for users with slow braking reaction, the second path may increase the frequency of emergency braking exercises; and for users with poor speed control, more scenarios involving constant speed driving may be introduced.

[0084] This cycle mechanism ensures the dynamic adaptation of the training path to the user's skills, so that the training is always scientific and effective. Through repeated path optimization and training, users can gradually solve their weak points and continuously improve their driving skills.

[0085] This step uses the real-time feedback and dynamic optimization technology of the newly added driving behavior data to achieve adaptive adjustment of the training path. Compared with the traditional fixed training method, this incremental learning and feedback mechanism significantly improves the personalization and accuracy of the training effect, allowing users to achieve the ultimate driving skill goal in progressive and efficient learning.

[0086] In summary, the embodiments of the present application have at least the following technical effects:

[0087] This application obtains virtual examination room scenes through a diversified virtual examination room construction system, and allows training users to log in to the system for driving training, outputs driving behavior data sets in real time for scoring, analyzes the data based on the driving score, and generates a personalized training path by combining historical data when the driving score does not meet the standard. The path consists of multiple training scenario nodes, and the training users perform step-by-step training according to the path until the preset skill score target is achieved.

[0088] It has achieved the technical effect of improving the personalization and pertinence of driving skills training, ensuring that the training content is highly matched with the students' actual skill needs, and improving the efficiency and effectiveness of training.

[0089] It should be noted that the above-mentioned sequence of the embodiments of the present application is only for description and does not represent the advantages and disadvantages of the embodiments. And the above-mentioned specific embodiments of this specification are described. In addition, the processes depicted in the accompanying drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0090] The above description is only a preferred embodiment of the present application and is not intended to limit the present application. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present application should be included in the protection scope of the present application.

[0091] This specification and the drawings are merely exemplary illustrations of the present application and are deemed to cover any and all modifications, variations, combinations or equivalents within the scope of the present application. Obviously, a person skilled in the art may make various modifications and variations to the present application without departing from the scope of the present application. Thus, if these modifications and variations of the present application fall within the scope of the present application and its equivalents, the present application intends to include these modifications and variations.

Claims

1. A driving skills training method based on a diversified virtual examination room, characterized in that: The method comprises: Based on the diversified virtual examination room construction system, obtain the virtual examination room scene; The training user logs in to the diversified virtual test site construction system, drives based on the virtual test site scene, and outputs a real-time driving behavior data set; Performing a driving score on the real-time driving behavior data set and outputting a driving score; If the driving score of the real-time driving behavior data set is less than a preset score, obtaining the historical driving behavior data set of the training user; Inputting the real-time driving behavior dataset and the historical driving behavior dataset into a driving path generation model, and acquiring a first driving training path according to the driving path generation model, wherein the first driving training path includes a plurality of path nodes, each path node corresponding to a driving training scenario; The training user performs driving training according to the first driving training path until the preset score is met.

2. The method according to claim 1, characterized in that The method for acquiring a first driving training path according to the driving path generation model includes: The driving path generation model extracts driving evaluation item features according to the real-time driving behavior data set and the historical driving behavior data set to obtain evaluation indicators corresponding to each driving evaluation item; Determining the skill items to be improved according to the evaluation indicators corresponding to the various driving evaluation items; Build a virtual examination scene library; The first driving training path is generated by performing scene matching in a virtual examination room scene library according to the skill item to be improved.

3. The method according to claim 2, characterized in that Construct a virtual examination room scene library, including: Set up multiple virtual exam scene samples; Collect driving behavior sample data of the training user group based on various virtual test site scene samples; Uploading the driving behavior sample data of each virtual test scene sample to the cloud platform, analyzing the driving behavior sample data of each virtual test scene sample on the cloud platform, and determining the skill improvement label of each virtual test scene sample; The multiple virtual examination scene samples are labeled according to the skill improvement tags to generate a virtual examination scene library.

4. The method according to claim 2, characterized in that According to the skill item to be improved, scene matching is performed in a virtual test site scene library to generate the first driving training path, the method comprising: According to the virtual test site scenario library, a plurality of driving training scenarios are obtained; Marking the difficulty of the plurality of driving training scenarios and outputting a plurality of difficulty labels, wherein the plurality of driving training scenarios correspond to the plurality of difficulty labels; According to the multiple difficulty labels, the multiple driving training scenarios are sequentially connected from small to large to generate the first driving training path.

5. The method according to claim 2, characterized in that The skill items to be improved are determined according to the evaluation indicators corresponding to the various driving evaluation items, wherein the various driving evaluation items at least include a braking evaluation item, a speed control evaluation item, a steering operation evaluation item, an environmental adaptability evaluation item, and an emergency response capability evaluation item.

6. The method according to claim 1, characterized in that After the training user performs driving training according to the first driving training path, the method further includes: When the training user performs driving training according to the first driving training path, recording a new driving behavior data set of the training user; Inputting the newly added driving behavior data set as incremental data into the driving path generation model for feedback learning, updating the first driving training path, and outputting a second driving path; The training user performs driving training according to the second driving training path.

7. The method according to claim 1, characterized in that The diversified virtual examination room construction system includes multiple construction components, and the multiple construction components include a scene type construction component, a difficulty level construction component, a scene task distribution component, and an environment simulation component; Collaborative construction is performed according to the multiple construction components to obtain a virtual examination room scene.

8. The method according to claim 1, characterized in that The diversified virtual examination room construction system includes a virtual reality operating device, and the training user performs a perceptual sensitivity test according to the virtual reality operating device to output a perceptual test data set; When the perception test data set meets the preset perception sensitivity, obtaining device configuration parameters; A connection is established between the virtual reality operation device and the training user according to the device configuration parameters.

Images