Bad driving behavior feedback intervention method and device based on driving behavior big data

Abstract

The present invention relates to the field of traffic safety technology, and specifically to a method and device for feedback intervention of bad driving behavior based on driving behavior big data. Through vehicle operation related data, driving environment data, driver behavior and physiological data. After pre-processing the original data, key features are extracted, a driving behavior feature database is constructed and a driving behavior pattern is generated. Bad driving behavior is determined by combining rule judgment, machine learning algorithm and pattern matching, wherein the rule judgment is based on historical big data statistical analysis, corresponding intervention measures are implemented for different types of bad driving behavior, and the driving behavior change data after intervention is recorded and transmitted back to the cloud to optimize subsequent strategies. The present invention integrates multiple technologies to achieve high-precision judgment and personalized feedback intervention of bad driving behavior, with the beneficial effects of real-time and high accuracy, personalized adaptation, multimodal feedback, privacy protection and data security, and can effectively improve driving safety.

Description

Technical Field

[0001] The present invention relates to the field of traffic safety technology, and in particular to a method and device for intervening in adverse driving behavior feedback based on driving behavior big data. Background Art

[0002] In today's transportation sector, with the continued rise in vehicle ownership, traffic safety is becoming increasingly severe. Traditional methods for monitoring driving behavior rely primarily on a single data source, such as tracking vehicle trajectories based solely on GPS data or conducting post-accident analysis based on relevant data. However, this approach has numerous drawbacks.

[0003] On the one hand, the monitoring dimension is extremely limited, only capturing a subset of driving information from a specific perspective, and failing to comprehensively and comprehensively consider the complex behaviors of drivers during actual driving. For example, relying solely on GPS makes it difficult to accurately determine whether a driver is distracted or fatigued, as it cannot capture the vehicle's internal operating conditions or the driver's physiological and behavioral status.

[0004] On the other hand, the timeliness of feedback is severely delayed. Data collection and analysis is usually conducted only after an accident has occurred, by which time the bad driving behavior has already led to serious consequences. It is impossible to provide timely warnings and corrections to the driver during driving, making it difficult to effectively prevent accidents.

[0005] Furthermore, intervention measures lack specificity and effectiveness. Because they cannot accurately identify the type and severity of unhealthy driving behaviors, intervention measures are often general and cannot be tailored to individual drivers and specific driving scenarios. This makes it difficult to truly leverage interventions and reduce the incidence of unhealthy driving behaviors.

[0006] In summary, the existing driving behavior monitoring and intervention methods can no longer meet the growing demand for traffic safety. There is an urgent need for an innovative system and method that can combine multi-source data and use advanced technologies to achieve comprehensive, real-time, and accurate monitoring and effective intervention of driving behavior, thereby improving road traffic safety assurance capabilities. Summary of the Invention

[0007] In view of the shortcomings of the current relevant existing technologies, the present invention provides a method and device for feedback intervention of bad driving behavior based on driving behavior big data. It uses multi-source driving behavior data to realize accurate real-time identification and classification of bad driving behavior, and provides personalized and diversified feedback based on the driver's historical records and preferences to help them correct it in time. It also dynamically optimizes the intervention strategy based on the behavior data and feedback effect, improves the effectiveness of intervention, reduces the incidence of traffic accidents, and ensures the safety of road users. It overcomes the shortcomings of existing technologies such as insufficient monitoring dimensions, untimely feedback, and lack of targeted intervention.

[0008] To achieve the above objectives, the present invention provides a first aspect of a method for providing feedback and intervention on bad driving behavior based on driving behavior big data, comprising the following steps:

[0009] Multi-source data collection steps: Use the vehicle's built-in data acquisition interface to collect vehicle operation-related data, use positioning and environmental perception equipment to collect driving environment data, and use image acquisition and wearable devices to collect driver behavior and physiological data;

[0010] Data processing and analysis steps: Preprocess the collected raw data to remove noise and outliers, extract key features including vehicle operating dynamics, driving environment conditions, driver operation and physiological status, aggregate a large amount of driving behavior data through a cloud platform to build a driving behavior feature database, and use data mining algorithms to generate driving behavior patterns;

[0011] Bad driving behavior determination steps: A combination of rule-based determination based on statistical analysis of historical data, machine learning algorithms using multi-source data as input, and pattern matching methods that build models based on the driver's historical behavior is used to determine whether the driving behavior is bad.

[0012] Personalized feedback steps: Provide feedback to the driver through one or more of visual, voice, and tactile feedback channels based on the type and severity of the adverse driving behavior, and dynamically adjust the intensity and form of the feedback based on the driver's historical driving data and behavioral tendencies;

[0013] Intervention strategy implementation and optimization steps: Implement corresponding intervention measures for different types of bad driving behaviors, and record the driving behavior change data after the intervention and transmit it back to the cloud for continuous optimization of subsequent intervention strategies.

[0014] Furthermore, in the multi-source data collection step, collecting vehicle operation related data includes collecting one or more data of vehicle speed, acceleration, braking depth, throttle depth, clutch depth, steering wheel angle, gear position, and engine speed through a vehicle bus;

[0015] Collecting driving environment data includes collecting one or more of real-time location, driving trajectory and road speed limit through GPS; obtaining surrounding environment data through external sensors;

[0016] Collecting driver behavior and physiological data includes using a camera to capture one or more behavioral characteristics of the driver's facial expression, head posture, and gaze direction; and using a wearable device to record the driver's physiological signals, including at least heart rate and fatigue.

[0017] Furthermore, in the data processing and analysis steps, the vehicle operation dynamics include at least the number of accelerations, the intensity of sudden braking and the lane deviation range; the driving environment conditions include at least the ratio of the current vehicle speed to the speed limit and the rate of change of the distance to the vehicle in front; the driver's operation and physiological state include at least one or more behavioral characteristics of whether the driver's line of sight deviates from the front, whether the facial expression shows fatigue, whether the driver lowers his head to shift gears and steers the wheel with one hand.

[0018] Furthermore, in the bad driving behavior determination step, the bad driving behavior classification includes at least distracted driving, fatigue driving, and dangerous driving; wherein the determination step is:

[0019] Rule determination: Based on historical big data statistical analysis, such as speeding judgment: current speed > road speed limit + 20, if this condition is met, it is judged as speeding and dangerous driving;

[0020] Rapid acceleration judgment: acceleration change in a short period of time> 3m / s 2 ,The time window is 2 seconds; Lane departure judgment: The vehicle deviates from the lane line by an angle greater than 15° and the duration is greater than 1.5 seconds;

[0021] Fatigue driving judgment: Facial state check shows eyes closed for more than 1.5 seconds and frequent dozing;

[0022] Machine learning model: Uses a random forest algorithm to classify bad driving behaviors. Input features include vehicle data, driving environment data, and driver behavior data. The model is trained using large-scale driving behavior data. A grid search is used to determine the optimal number, depth, and splitting criteria of trees, determine the driving behavior category, and output the probability value of bad behavior.

[0023] Behavior pattern matching: Build a behavior model based on the driver's historical data, match real-time behavior with historical behavior patterns, and identify abnormal behavior. Specifically, collect the driver's long-term historical driving data and build a driving behavior model based on time series features. Collect speed change patterns, sudden acceleration and braking frequency, and lane keeping stability features, use a time series model (LSTM) to extract the temporal dependency of driving behavior, match real-time driving behavior data with historical behavior data, and calculate the similarity: 1-|Current Behavior Pattern - Historical Behavior Pattern|>0.8, then identify abnormal behavior.

[0024] Furthermore, in the data processing and analysis steps, the data aggregation and storage operations are as follows:

[0025] Individual feature database construction: For each driver, a personalized model is created by recording and analyzing their long-term personal driving data, covering both normal and abnormal behavior patterns;

[0026] Global feature database construction: Aggregating large-scale driver behavior data, refining the behavioral patterns and distribution characteristics of all drivers, using this as the core of a global reference benchmark to generate a baseline model for judgment. This includes the average frequency distribution of sudden braking by drivers in different road scenarios, and information on common driving behavior differences in specific traffic environments in urban or rural areas.

[0027] The data mining and behavior pattern generation process is as follows:

[0028] Generate regular behavior patterns: Statistical analysis and clustering algorithms are used to process historical driver data. The data is input with characteristics such as speed range, braking force, acceleration change, and lane keeping deviation amplitude. The data is first cleaned and normalized to remove outliers, followed by feature extraction and dimensionality reduction. Finally, a clustering algorithm is run to classify the behavior data, outputting a personalized regular driving behavior template for each driver, including speed-time distribution, braking force range, and steering frequency.

[0029] Abnormal behavior pattern generation: Association rule algorithms and anomaly detection algorithms are used to mine potential abnormal patterns. A driving behavior feature vector is constructed, including elements such as sudden acceleration amplitude, lane deviation angle, and speeding duration. The association rule algorithm is used to analyze the combination of behavioral features. The isolation forest algorithm is used to detect anomalies in driving behavior data, marking behavior patterns that deviate from the global feature distribution. The abnormal behavior rule table is output, covering abnormal behavior feature combinations and their corresponding probabilities.

[0030] Furthermore, the specific implementation details of the big data processing method are as follows:

[0031] Data processing steps and algorithms:

[0032] Data cleaning: Use interpolation methods based on mean replacement or time series interpolation to clean up missing values, outliers, and sensor noise;

[0033] Feature extraction: Extracting core driving features from raw data, including vehicle features, environmental features, and driver features;

[0034] Feature normalization: Min-Max normalization is used to scale the data to the [0,1] interval;

[0035] Feature storage: The cleaned and normalized feature data is stored in individual and global databases according to the hierarchical structure;

[0036] Output data format:

[0037] The data is stored in a multidimensional feature table, where each row corresponds to a driving behavior record;

[0038] The generated rules are stored as a rule condition table, which describes the conditions and rules of the behavior pattern in detail;

[0039] The processed data is applied as follows:

[0040] Supports personalized judgment: Combined with the individual feature database, judgment criteria are tailored to the driver's driving habits. For example, if a driver's sudden braking force threshold is higher than the group average, the system can dynamically adapt the abnormal judgment criteria to him.

[0041] Optimize the global model: The global feature database provides basic rules and behavioral baselines for the judgment module. New driving data is regularly integrated into the global database to optimize model parameters and improve judgment accuracy.

[0042] A second aspect of the present invention provides a device for providing feedback and intervention on bad driving behavior based on driving behavior big data, the device being used to implement the above-mentioned method, comprising:

[0043] Data acquisition unit: used to perform multi-source data acquisition steps, using the vehicle's built-in data acquisition interface to collect vehicle operation-related data, using positioning and environmental perception equipment to collect driving environment data, and using image acquisition and wearable devices to collect driver behavior and physiological data;

[0044] Data processing and analysis unit: This unit pre-processes the raw data collected by the acquisition unit to remove noise and outliers, extracts key features including vehicle operating dynamics, driving environment conditions, driver operation, and physiological status, aggregates large amounts of driving behavior data through a cloud platform to build a driving behavior feature database, and uses data mining algorithms to generate driving behavior patterns.

[0045] Bad Driving Behavior Determination Unit: This unit uses a combination of rule-based determination based on statistical analysis of historical data, a machine learning algorithm with multi-source data as input, and a pattern matching method that builds a model based on the driver's historical behavior to determine whether driving behavior is bad.

[0046] Personalized feedback unit: Provides feedback to the driver through one or more of visual, voice, and tactile feedback channels based on the type and severity of adverse driving behavior, and dynamically adjusts the intensity and form of feedback based on the driver's historical driving data and behavioral tendencies;

[0047] Intervention strategy implementation unit: Implements corresponding intervention measures for different types of bad driving behaviors, and records the driving behavior change data after the intervention and transmits it back to the cloud for continuous optimization of subsequent intervention strategies;

[0048] System architecture unit: includes vehicle-side devices, cloud platform and edge computing unit. The vehicle-side devices deploy real-time judgment and feedback functions to ensure the real-time and high efficiency of the system; the cloud platform conducts large-scale data training and analysis to optimize model parameters; the edge computing unit realizes real-time data preprocessing and preliminary judgment through the on-board computing unit.

[0049] Furthermore, the data acquisition unit includes:

[0050] Vehicle data acquisition subunit: collects one or more data including vehicle speed, acceleration, braking depth, throttle depth, clutch depth, steering wheel angle, gear position and engine speed through the vehicle bus;

[0051] Driving environment data acquisition subunit: collects one or more of real-time location, driving trajectory and road speed limit through GPS, and obtains surrounding environment data through external sensors;

[0052] Driver data acquisition subunit: The camera captures one or more behavioral characteristics of the driver's facial expression, head posture and gaze direction, and the wearable device records the driver's physiological signals, including at least heart rate and fatigue.

[0053] Furthermore, the data processing and analysis unit includes:

[0054] Data cleaning and preprocessing subunit: used to clean up missing values, outliers and sensor noise in collected data;

[0055] Feature extraction subunit: extracts vehicle features from the raw data, such as sudden braking intensity, acceleration changes, and steering wheel rotation angle; environmental features, such as the ratio of vehicle speed to speed limit and the rate of change of the distance to the preceding vehicle; and driver features, such as the driver's head posture and line of sight.

[0056] Big data processing sub-unit: Aggregates data through the cloud platform to build a driving behavior feature database, and uses data mining algorithms to generate driving behavior patterns. This includes using statistical analysis and clustering algorithms to generate regular behavior patterns, and using association rule algorithms and anomaly detection algorithms to generate abnormal behavior patterns. The data is then processed using Min-Max standardization and stored in a multidimensional feature table and rule condition table.

[0057] Furthermore, in the system architecture unit:

[0058] The vehicle-side equipment includes an on-board processor and a storage device. The on-board processor is used to run real-time judgment and feedback programs, and the storage device is used to store temporary data and some key algorithms.

[0059] The cloud platform includes large-scale data storage servers and high-performance computing clusters for storing massive amounts of driving behavior data and conducting complex data training and analysis;

[0060] The edge computing unit includes an on-board edge computing chip and related software modules to realize real-time data preprocessing and preliminary judgment functions, and data is transmitted between the vehicle-side equipment, cloud platform and edge computing unit through a secure communication network.

[0061] By integrating rule determination, machine learning models, and behavioral pattern matching, the present invention achieves high-precision determination and personalized feedback intervention for adverse driving behaviors, with the following beneficial effects:

[0062] 1. Real-time and high accuracy: Combining cloud-edge collaborative computing with machine learning algorithms to achieve fast and efficient behavior judgment.

[0063] 2. Personalized adaptation: Through behavioral pattern matching and adaptive learning, system parameters are dynamically adjusted to adapt to the habits of different drivers.

[0064] 3. Multimodal feedback: Provide feedback through voice, vision, touch, and other means to improve the effectiveness of intervention.

[0065] 4. Privacy protection and data security: Federated learning and data encryption technologies are used to ensure driver data security and privacy. BRIEF DESCRIPTION OF THE DRAWINGS

[0066] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. 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 paying any creative work.

[0067] Figure 1 It is a flow chart of the driver safety awareness assessment and optimization method of the present invention. DETAILED DESCRIPTION

[0068] Exemplary embodiments will be described in detail herein, examples of which are illustrated in the accompanying drawings. In the following description, when referring to the drawings, like numbers in different figures represent like or similar elements unless otherwise indicated. The embodiments described in the following exemplary embodiments are not intended to represent all possible embodiments consistent with the present invention. Rather, they are merely examples of apparatus and methods consistent with certain aspects of the present invention, as detailed in the appended claims.

[0069] like Figure 1As shown, the first aspect of this embodiment provides a method for feedback intervention of bad driving behavior based on driving behavior big data, comprising the following steps:

[0070] Multi-source data collection steps: Use the vehicle's built-in data acquisition interface to collect vehicle operation-related data, use positioning and environmental perception equipment to collect driving environment data, and use image acquisition and wearable devices to collect driver behavior and physiological data;

[0071] Data processing and analysis steps: Preprocess the collected raw data to remove noise and outliers, extract key features including vehicle operating dynamics, driving environment conditions, driver operation and physiological status, aggregate a large amount of driving behavior data through a cloud platform to build a driving behavior feature database, and use data mining algorithms to generate driving behavior patterns;

[0072] Bad driving behavior determination steps: A combination of rule-based determination based on statistical analysis of historical data, machine learning algorithms using multi-source data as input, and pattern matching methods that build models based on the driver's historical behavior is used to determine whether the driving behavior is bad.

[0073] Personalized feedback steps: Provide feedback to the driver through one or more of visual, voice, and tactile feedback channels based on the type and severity of the adverse driving behavior, and dynamically adjust the intensity and form of the feedback based on the driver's historical driving data and behavioral tendencies;

[0074] Intervention strategy implementation and optimization steps: Implement corresponding intervention measures for different types of bad driving behaviors, and record the driving behavior change data after the intervention and transmit it back to the cloud for continuous optimization of subsequent intervention strategies.

[0075] As a preferred embodiment, in the multi-source data collection step of this embodiment, collecting vehicle operation related data includes collecting one or more data of vehicle speed, acceleration, braking depth, throttle depth, clutch depth, steering wheel angle, gear position and engine speed through the vehicle bus;

[0076] Collecting driving environment data includes collecting one or more of real-time location, driving trajectory and road speed limit through GPS; obtaining surrounding environment data through external sensors;

[0077] Collecting driver behavior and physiological data includes using a camera to capture one or more behavioral characteristics of the driver's facial expression, head posture, and gaze direction; and using a wearable device to record the driver's physiological signals, including at least heart rate and fatigue.

[0078] As a preferred implementation, in the data processing and analysis steps of this embodiment, the vehicle operation dynamics include at least the number of accelerations, the intensity of sudden braking, and the lane deviation range; the driving environment conditions include at least the ratio of the current vehicle speed to the speed limit, and the rate of change of the distance to the vehicle in front; the driver's operation and physiological state include at least one or more behavioral characteristics of whether the driver's line of sight deviates from the front, whether the facial expression shows fatigue, whether the driver lowers his head to shift gears, and whether he steers the wheel with one hand.

[0079] As a preferred implementation, in the bad driving behavior determination step in this embodiment, the bad driving behavior classification includes at least distracted driving, such as not looking ahead for a long time, operating a mobile phone, etc.

[0080] Fatigue driving, such as frequent closing of eyes, slow reaction, frequent dozing off, etc.

[0081] and dangerous driving, such as speeding, sudden acceleration, sudden braking, changing lanes without using turn signals, and not wearing seat belts.

[0082] The judgment steps are:

[0083] Rule determination: Based on historical big data statistical analysis, such as speeding judgment: current speed > road speed limit + 20, if this condition is met, it is judged as speeding and dangerous driving;

[0084] Rapid acceleration judgment: acceleration change in a short period of time> 3m / s 2 ,The time window is 2 seconds; Lane departure judgment: The vehicle deviates from the lane line by an angle greater than 15° and the duration is greater than 1.5 seconds;

[0085] Fatigue driving judgment: Facial state check shows eyes closed for more than 1.5 seconds and frequent dozing;

[0086] Machine learning model: Uses a random forest algorithm to classify bad driving behaviors. Input features include vehicle data, driving environment data, and driver behavior data. The model is trained using large-scale driving behavior data. A grid search is used to determine the optimal number, depth, and splitting criteria of trees, determine the driving behavior category, and output the probability value of bad behavior.

[0087] Behavior pattern matching: Build a behavior model based on the driver's historical data, match real-time behavior with historical behavior patterns, and identify abnormal behavior. Specifically, collect the driver's long-term historical driving data and build a driving behavior model based on time series features. Collect speed change patterns, sudden acceleration and braking frequency, and lane keeping stability features, use a time series model (LSTM) to extract the temporal dependency of driving behavior, match real-time driving behavior data with historical behavior data, and calculate the similarity: 1-|Current Behavior Pattern - Historical Behavior Pattern|>0.8, then identify abnormal behavior.

[0088] Personalized module feedback;

[0089] Dynamic feedback mechanism;

[0090] Provide various feedback methods based on the severity of the bad behavior:

[0091] Visual feedback: Specific warning information (e.g., “You are speeding 20%”) is presented on the vehicle’s display screen.

[0092] Voice Announcement: Play voice warnings, such as "Please stay focused."

[0093] Haptic feedback: Vibrates the steering wheel or seat to alert the driver to potential hazards.

[0094] Personalized adjustments;

[0095] The intensity and form of feedback can be adjusted based on the driver's driving history and behavioral preferences. For example, the frequency of tactile feedback can be increased for drivers with more fatigued driving habits.

[0096] intervention strategy module;

[0097] intervention program design;

[0098] Different intervention strategies are adopted for different bad driving behaviors:

[0099] Distracted driving: The system plays a warning tone in real time and pauses the entertainment system (such as music playback).

[0100] Fatigue driving: prompts the driver to take a break and is accompanied by a visual warning.

[0101] Dangerous driving: Reduce vehicle speed by braking gently.

[0102] Evaluation of intervention effectiveness;

[0103] Record changes in driving behavior after each intervention and upload the data to the cloud to optimize subsequent intervention strategies.

[0104] As a preferred embodiment, in the data processing and analysis steps of this embodiment, the data aggregation and storage operations are as follows:

[0105] Construction of an individual characteristic database: For each driver, an individual database is formed into a personalized model through long-term recording and analysis of personal driving data. This includes normal behavior patterns (recording the driver's typical behavioral characteristics, such as speed distribution range, braking frequency, throttle response time, etc.) and abnormal behavior patterns (setting personalized abnormal behavior thresholds based on personal historical data, for example, a driver's speeding behavior occurs when the speed limit is exceeded by +30, rather than the universal +20 standard).

[0106] Global feature database construction: Based on large-scale driver behavior data, the behavioral patterns and distribution characteristics of all drivers are extracted. The core function of the global database is to provide a global reference standard and form a baseline model for judgment. For example, the average frequency of sudden braking of drivers in different road scenarios and the common driving behavior differences in regional traffic environments (urban and rural roads) are analyzed.

[0107] The data mining and behavior pattern generation process is as follows: through a variety of algorithms and analysis steps, the driver's normal behavior patterns and abnormal behavior patterns are generated to support the subsequent bad driving behavior judgment module.

[0108] Normal behavior pattern:

[0109] Algorithm: Statistical analysis and clustering algorithms (such as K-Means clustering) are used to cluster drivers' historical driving data. Input features include speed range, braking force, acceleration changes, lane keeping deviation amplitude, etc.

[0110] step:

[0111] Clean and normalize driving data to remove outliers (such as sensor noise).

[0112] Extract driving data features and perform data dimensionality reduction (such as PCA).

[0113] Run a clustering algorithm to divide the driver's behavior data into different categories (such as normal driving, mild speeding, high-frequency sudden braking, etc.).

[0114] Output data: Each driver's personalized regular driving behavior template, including speed-time distribution, braking force range, steering frequency, etc.

[0115] Abnormal behavior patterns

[0116] Algorithms: Use association rule algorithms and anomaly detection algorithms to identify potential abnormal patterns in driving behavior.

[0117] Association rule algorithm: used to analyze the combination relationship between behavioral features, such as whether sudden braking and lane departure occur at the same time.

[0118] Anomaly detection algorithm: Use the Isolation Forest algorithm to extract behavioral features that do not conform to normal patterns from the data.

[0119] Steps: Construct a driving behavior feature vector (such as rapid acceleration amplitude, lane deviation angle, and speeding duration).

[0120] Apply association rule algorithms to analyze the combination of behavioral features (such as whether speeding is accompanied by sudden braking).

[0121] Perform outlier detection on driving behavior data to mark behavioral patterns that deviate from the global feature distribution.

[0122] Output data: abnormal behavior rule table, including abnormal behavior feature combinations (such as "sudden acceleration + lane departure") and their corresponding probabilities.

[0123] As a preferred implementation, this embodiment also includes a method for processing big data, and the specific implementation details are as follows:

[0124] Data processing steps and algorithms:

[0125] Data cleaning: Cleans missing values, outliers, and sensor noise. Algorithms include interpolation methods based on mean substitution or time series interpolation.

[0126] Feature extraction: Extracting core driving features from raw data, such as:

[0127] Vehicle characteristics: sudden braking intensity, acceleration changes, steering wheel rotation angle, etc.

[0128] Environmental characteristics: ratio of vehicle speed to speed limit, rate of change of distance to the preceding vehicle, etc.

[0129] Driver characteristics: driver's head posture, gaze direction, etc.

[0130] Feature normalization: Min-Max normalization is used to scale the data to the range of [0, 1].

[0131] Feature storage: The cleaned and normalized feature data is stored hierarchically in individual and global databases.

[0132] Output data format:

[0133] The data is stored as a multidimensional feature table (in Alibaba Cloud's AnalysisDB database), where each row represents a driving behavior record.

[0134] The generated rules are stored as rule condition tables, which describe the conditions and rules of the behavior pattern (such as "IF rapid acceleration > 3m / s 2 AND braking force > X THEN it is determined to be abnormal behavior").

[0135] The processed data is applied as follows:

[0136] Support personalized judgment: Combined with the individual feature database, it can provide drivers with personalized judgment standards that are more in line with their driving habits.

[0137] For example, if a driver's force threshold when braking suddenly is usually higher than the group average, the system can dynamically adjust its abnormal judgment criteria.

[0138] Optimize the global model: The global feature database provides basic rules and behavioral baselines for the judgment module (such as the group speeding judgment threshold).

[0139] Regularly integrate new driving data into the global database to optimize model parameters and improve judgment accuracy.

[0140] A second aspect of the present invention provides a device for providing feedback and intervention on bad driving behavior based on driving behavior big data, the device being used to implement the above-mentioned method, comprising:

[0141] Data acquisition unit: used to perform multi-source data acquisition steps, using the vehicle's built-in data acquisition interface to collect vehicle operation-related data, using positioning and environmental perception equipment to collect driving environment data, and using image acquisition and wearable devices to collect driver behavior and physiological data;

[0142] Data processing and analysis unit: This unit pre-processes the raw data collected by the acquisition unit to remove noise and outliers, extracts key features including vehicle operating dynamics, driving environment conditions, driver operation, and physiological status, aggregates large amounts of driving behavior data through a cloud platform to build a driving behavior feature database, and uses data mining algorithms to generate driving behavior patterns.

[0143] Bad Driving Behavior Determination Unit: This unit uses a combination of rule-based determination based on statistical analysis of historical data, a machine learning algorithm with multi-source data as input, and a pattern matching method that builds a model based on the driver's historical behavior to determine whether driving behavior is bad.

[0144] Personalized feedback unit: Provides feedback to the driver through one or more of visual, voice, and tactile feedback channels based on the type and severity of adverse driving behavior, and dynamically adjusts the intensity and form of feedback based on the driver's historical driving data and behavioral tendencies;

[0145] Intervention strategy implementation unit: Implements corresponding intervention measures for different types of bad driving behaviors, and records the driving behavior change data after the intervention and transmits it back to the cloud for continuous optimization of subsequent intervention strategies;

[0146] System architecture unit: includes vehicle-side devices, cloud platform and edge computing unit. The vehicle-side devices deploy real-time judgment and feedback functions to ensure the real-time and high efficiency of the system; the cloud platform conducts large-scale data training and analysis to optimize model parameters; the edge computing unit realizes real-time data preprocessing and preliminary judgment through the on-board computing unit.

[0147] Furthermore, the data acquisition unit includes:

[0148] Vehicle data acquisition subunit: collects one or more data including vehicle speed, acceleration, braking depth, throttle depth, clutch depth, steering wheel angle, gear position and engine speed through the vehicle bus;

[0149] Driving environment data acquisition subunit: collects one or more of real-time location, driving trajectory and road speed limit through GPS, and obtains surrounding environment data through external sensors;

[0150] Driver data acquisition subunit: The camera captures one or more behavioral characteristics of the driver's facial expression, head posture and gaze direction, and the wearable device records the driver's physiological signals, including at least heart rate and fatigue.

[0151] Furthermore, the data processing and analysis unit includes:

[0152] Data cleaning and preprocessing subunit: used to clean up missing values, outliers and sensor noise in collected data;

[0153] Feature extraction subunit: extracts vehicle features from the raw data, such as sudden braking intensity, acceleration changes, and steering wheel rotation angle; environmental features, such as the ratio of vehicle speed to speed limit and the rate of change of the distance to the preceding vehicle; and driver features, such as the driver's head posture and line of sight.

[0154] Big data processing sub-unit: Aggregates data through the cloud platform to build a driving behavior feature database, and uses data mining algorithms to generate driving behavior patterns. This includes using statistical analysis and clustering algorithms to generate regular behavior patterns, and using association rule algorithms and anomaly detection algorithms to generate abnormal behavior patterns. The data is then processed using Min-Max standardization and stored in a multidimensional feature table and rule condition table.

[0155] Furthermore, in the system architecture unit:

[0156] The vehicle-side equipment includes an on-board processor and a storage device. The on-board processor is used to run real-time judgment and feedback programs, and the storage device is used to store temporary data and some key algorithms.

[0157] The cloud platform includes large-scale data storage servers and high-performance computing clusters for storing massive amounts of driving behavior data and conducting complex data training and analysis;

[0158] The edge computing unit includes an on-board edge computing chip and related software modules to realize real-time data preprocessing and preliminary judgment functions, and data is transmitted between the vehicle-side equipment, cloud platform and edge computing unit through a secure communication network.

[0159] In the field of driving behavior monitoring and intervention, the present invention brings significant improvements in many aspects.

[0160] In terms of accuracy and timeliness, the system significantly enhances its ability to identify adverse driving behavior by integrating rule-based decision-making, machine learning models, and behavioral pattern matching. Rule-based decision-making relies on precise criteria derived from statistical analysis of massive amounts of historical big data, such as clear, quantitative criteria for speeding, rapid acceleration, lane departure, and fatigued driving. This allows it to quickly identify obvious signs of adverse driving behavior. The machine learning model, leveraging the Random Forest algorithm, leverages its powerful classification capabilities. In-depth training is conducted on a rich and diverse set of vehicle, driving environment, and driver behavior data, enabling it to handle complex driving scenarios and behavior patterns. The outputted adverse behavior probability provides a reliable reference for decision-making, effectively overcoming the limitations of single rule-based decision-making. Behavioral pattern matching leverages long-term accumulated historical driver data and employs a time series model (LSTM) to extract the temporal dependencies of driving behavior. Through meticulous matching and similarity calculation with real-time behavior, it can discern subtle anomalies in driver behavior, making even uncommon or unique adverse driving behaviors difficult to escape unnoticed. This multi-dimensional, integrated decision-making mechanism ensures that the system can quickly and accurately identify adverse driving behavior in a variety of driving situations, laying a solid foundation for timely subsequent intervention.

[0161] In terms of personalized services, the present invention demonstrates excellent adaptability. The individual feature database constructed for each driver records in detail the unique behavioral patterns of their personal driving process, including conventional behavioral characteristics such as speed distribution range, braking frequency, throttle response time, and personalized abnormal behavior thresholds set based on personal historical data. For example, the normal speed range of some drivers under specific road conditions may be different from the general standard. The system can make targeted judgments and feedback based on this, avoiding misjudgments or inappropriate feedback caused by unified standards. At the same time, based on the driver's historical driving records and behavioral preferences, the personalized feedback unit can accurately adjust the intensity and form when providing feedback. For drivers who are prone to fatigue, increasing the frequency of tactile feedback or adjusting the tone and content of voice prompts will make it easier for them to accept and pay attention to feedback information, thereby more effectively helping drivers correct bad behaviors, improving driving safety, and enhancing the interactive experience and trust between the driver and the system.

[0162] The application of multimodal feedback mechanisms significantly enhances the effectiveness of intervention. Visual feedback uses the on-board display to intuitively display specific and eye-catching warning messages such as "You are speeding by [X]%" and "Lane deviation, please adjust." This allows drivers to immediately understand the type and severity of their driving misconduct and respond quickly. Voice notification uses clear and distinct warnings such as "Please stay focused, the road ahead is complex" and "You are driving fatigued, please rest as soon as possible." These messages convey important reminders without disrupting the driver's vision, enhancing the timeliness and effectiveness of feedback. Haptic feedback uses vibrations in the steering wheel or seat to provide the driver with strong physical stimulation when critical or critical driving misconduct occurs, such as impending collision or excessive speeding. This quickly alerts the driver and prompts them to take corrective action. The three feedback methods complement each other, comprehensively covering the driver's perceptual range, effectively increasing the driver's attention to and willingness to correct their misconduct, and significantly reducing the risk of traffic accidents caused by misconduct.

[0163] At the data processing and system architecture level, the data acquisition and feature extraction module comprehensively collects data from multiple sources, covering key vehicle operating parameters, detailed information about the driving environment, and the driver's behavior and physiological state, providing rich and comprehensive data support for the system. Big data processing technologies efficiently aggregate, store, and deeply mine this massive amount of data. The resulting driving behavior feature database not only includes personalized models for individual drivers but also extracts behavioral patterns and distribution characteristics for the entire population, providing strong data support and decision-making basis for the precise operation of the entire system. The cloud-vehicle collaborative system architecture design fully leverages the advantages of all parties. The real-time judgment and feedback functions of the vehicle-side equipment ensure the system's immediate response during driving, meeting real-time requirements. The cloud platform, with its powerful computing power and large-scale data storage capabilities, conducts in-depth data training and analysis, continuously optimizes model parameters, and improves overall system performance and accuracy. The edge computing unit, through the on-board computing unit, performs real-time data preprocessing and preliminary judgment, reducing the burden on both the cloud and vehicle sides, improving data processing efficiency, and ensuring smooth system operation.

[0164] In terms of privacy protection, encryption technology is used to encrypt uploaded data, effectively preventing theft or tampering during transmission and ensuring the confidentiality of driver data. The application of federated learning technology is a major innovation, enabling joint optimization of decision models without sharing original data. This fully leverages the value of multi-source data while maximizing driver privacy and security, allowing drivers to use this system with peace of mind and enhancing the system's credibility and acceptability.

[0165] The present invention has significant advantages in monitoring, feedback and intervention of bad driving behavior, can effectively improve driving safety, and provides an efficient, intelligent and reliable solution for road traffic safety management.

[0166] Although the embodiments of the present invention have been shown and described above, it will be understood that the above embodiments are exemplary and are not to be construed as limitations on the present invention. A person skilled in the art may change, modify, replace and modify the above embodiments within the scope of the present invention.

Claims

1. A method for feedback intervention of bad driving behavior based on driving behavior big data, characterized by: The following steps are involved: Multi-source data collection steps: Use the vehicle's built-in data acquisition interface to collect vehicle operation-related data, use positioning and environmental perception equipment to collect driving environment data, and use image acquisition and wearable devices to collect driver behavior and physiological data; Data processing and analysis steps: Preprocess the collected raw data to remove noise and outliers, extract key features including vehicle operating dynamics, driving environment conditions, driver operation and physiological status, aggregate a large amount of driving behavior data through a cloud platform to build a driving behavior feature database, and use data mining algorithms to generate driving behavior patterns; Bad driving behavior determination steps: A combination of rule-based determination based on statistical analysis of historical data, machine learning algorithms using multi-source data as input, and pattern matching methods that build models based on the driver's historical behavior is used to determine whether the driving behavior is bad. Personalized feedback steps: Provide feedback to the driver through one or more of visual, voice, and tactile feedback channels based on the type and severity of the adverse driving behavior, and dynamically adjust the intensity and form of the feedback based on the driver's historical driving data and behavioral tendencies; Intervention strategy implementation and optimization steps: Implement corresponding intervention measures for different types of bad driving behaviors, and record the driving behavior change data after the intervention and transmit it back to the cloud for continuous optimization of subsequent intervention strategies; In the bad driving behavior determination step, the bad driving behavior classification includes at least distracted driving, fatigue driving, and dangerous driving; wherein the determination step is: Rule determination: Based on historical big data statistical analysis, including speeding judgment: current speed > road speed limit + 20, if this condition is met, it is judged as speeding and dangerous driving; Rapid acceleration judgment: acceleration change > 3m / s² in a short period of time, with a time window of 2 seconds; lane departure judgment: the vehicle's deviation angle from the lane line is > 15° and the duration is > 1.5 seconds; Fatigue driving judgment: Facial state check shows eyes closed for more than 1.5 seconds and frequent dozing; Machine learning model: Uses a random forest algorithm to classify bad driving behaviors. Input features include vehicle data, driving environment data, and driver behavior data. The model is trained using large-scale driving behavior data. A grid search is used to determine the optimal number, depth, and splitting criteria of trees, determine the driving behavior category, and output the probability value of bad behavior. Behavior pattern matching: Build a behavior model based on the driver's historical data, match real-time behavior with historical behavior patterns, and determine abnormal behavior. Specifically, collect the driver's long-term historical driving data and build a driving behavior model based on time series characteristics. Collect speed change patterns, sudden acceleration and braking frequency, and lane keeping stability characteristics, use a time series model to extract the time dependency of driving behavior, match real-time driving behavior data with historical behavior data, and calculate the similarity: 1-|Current behavior pattern-Historical behavior pattern|>0.8, then it is determined to be abnormal behavior.

2. The method according to claim 1, wherein: In the multi-source data collection step, collecting vehicle operation related data includes collecting one or more data of vehicle speed, acceleration, braking depth, throttle depth, clutch depth, steering wheel angle, gear position and engine speed through a vehicle bus; Collecting driving environment data includes collecting one or more of real-time location, driving trajectory and road speed limit through GPS; obtaining surrounding environment data through external sensors; Collecting driver behavior and physiological data includes capturing one or more behavioral characteristics of the driver's facial expression, head posture, and gaze direction through a camera; Wearable devices record the driver's physiological signals, including at least heart rate and fatigue.

3. The method according to claim 1, wherein: In the data processing and analysis steps, the vehicle operation dynamics include at least the number of accelerations, the intensity of sudden braking and the lane deviation range; the driving environment conditions include at least the ratio of the current vehicle speed to the speed limit and the rate of change of the distance to the vehicle in front; the driver's operation and physiological state include at least one or more behavioral characteristics of whether the driver's line of sight deviates from the front, whether the facial expression shows fatigue, whether the driver lowers his head to shift gears and steers the wheel with one hand.

4. The method according to any one of claims 1 to 3, characterized in that: The data processing and analysis steps include data aggregation and storage operations as follows: Individual feature database construction: For each driver, a personalized model is created by recording and analyzing their long-term personal driving data, covering both normal and abnormal behavior patterns; Global feature database construction: Aggregating large-scale driver behavior data, refining the behavioral patterns and distribution characteristics of all drivers, using this as the core of a global reference benchmark to generate a baseline model for judgment. This includes the average frequency distribution of sudden braking by drivers in different road scenarios, and information on common driving behavior differences in specific traffic environments in urban or rural areas. The data mining and behavior pattern generation process is as follows: Generate regular behavior patterns: Statistical analysis and clustering algorithms are used to process historical driver data. Input features include speed range, braking force, acceleration changes, and lane keeping deviation amplitude. The data is first cleaned and normalized to remove outliers. Feature extraction and dimensionality reduction are then performed. Finally, a clustering algorithm is run to categorize the behavior data. The resulting template, personalized for each driver, includes speed-time distribution, braking force range, and steering frequency. Abnormal behavior pattern generation: Association rule algorithms and anomaly detection algorithms are used to mine potential abnormal patterns. A driving behavior feature vector is constructed, including elements such as sudden acceleration amplitude, lane deviation angle, and speeding duration. The association rule algorithm is used to analyze the combination of behavioral features. The isolation forest algorithm is used to detect anomalies in driving behavior data, marking behavior patterns that deviate from the global feature distribution. The abnormal behavior rule table is output, covering abnormal behavior feature combinations and their corresponding probabilities.

5. The method according to claim 4, characterized in that: The specific implementation details of the big data processing method are as follows: Data processing steps and algorithms: Data cleaning: Use interpolation methods based on mean replacement or time series interpolation to clean up missing values, outliers, and sensor noise; Feature extraction: Extracting core driving features from raw data, including vehicle features, environmental features, and driver features; Feature normalization: Min-Max normalization is used to scale the data to the [0,1] interval; Feature storage: The cleaned and normalized feature data is stored in individual and global databases according to the hierarchical structure; Output data format: The data is stored in a multidimensional feature table, where each row corresponds to a driving behavior record; The generated rules are stored as a rule condition table, which describes the conditions and rules of the behavior pattern in detail; The processed data is applied as follows: Supports personalized judgment: Combined with the individual feature database, judgment criteria are tailored to the driver's driving habits. For example, if a driver's sudden braking force threshold is higher than the group average, the system can dynamically adapt the abnormal judgment criteria to him. Optimize the global model: The global feature database provides basic rules and behavioral baselines for the judgment module. New driving data is regularly integrated into the global database to optimize model parameters and improve judgment accuracy.

6. A bad driving behavior feedback intervention device based on driving behavior big data, characterized in that: The device is used to implement the method according to any one of claims 1 to 5, comprising: Data acquisition unit: used to perform multi-source data acquisition steps, using the vehicle's built-in data acquisition interface to collect vehicle operation-related data, using positioning and environmental perception equipment to collect driving environment data, and using image acquisition and wearable devices to collect driver behavior and physiological data; Data processing and analysis unit: This unit pre-processes the raw data collected by the acquisition unit to remove noise and outliers, extracts key features including vehicle operating dynamics, driving environment conditions, driver operation, and physiological status, aggregates large amounts of driving behavior data through a cloud platform to build a driving behavior feature database, and uses data mining algorithms to generate driving behavior patterns. Bad Driving Behavior Determination Unit: This unit uses a combination of rule-based determination based on statistical analysis of historical data, a machine learning algorithm with multi-source data as input, and a pattern matching method that builds a model based on the driver's historical behavior to determine whether driving behavior is bad. Personalized feedback unit: Provides feedback to the driver through one or more of visual, voice, and tactile feedback channels based on the type and severity of adverse driving behavior, and dynamically adjusts the intensity and form of feedback based on the driver's historical driving data and behavioral tendencies; Intervention strategy implementation unit: Implements corresponding intervention measures for different types of bad driving behaviors, and records the driving behavior change data after the intervention and transmits it back to the cloud for continuous optimization of subsequent intervention strategies; System architecture unit: includes vehicle-side devices, cloud platform and edge computing unit. The vehicle-side devices deploy real-time judgment and feedback functions to ensure the real-time and high efficiency of the system; the cloud platform conducts large-scale data training and analysis to optimize model parameters; the edge computing unit realizes real-time data preprocessing and preliminary judgment through the on-board computing unit.

7. The device according to claim 6, characterized in that The data acquisition unit includes: Vehicle data acquisition subunit: collects one or more data including vehicle speed, acceleration, braking depth, throttle depth, clutch depth, steering wheel angle, gear position and engine speed through the vehicle bus; Driving environment data acquisition subunit: collects one or more of real-time location, driving trajectory and road speed limit through GPS, and obtains surrounding environment data through external sensors; Driver data acquisition subunit: The camera captures one or more behavioral characteristics of the driver's facial expression, head posture and gaze direction, and the wearable device records the driver's physiological signals, including at least heart rate and fatigue.

8. The device according to claim 7, characterized in that The data processing and analysis unit includes: Data cleaning and preprocessing subunit: used to clean up missing values, outliers and sensor noise in collected data; Feature extraction subunit: extracts vehicle features from the raw data, including braking intensity, acceleration changes, and steering wheel angle; environmental features, including the ratio of vehicle speed to speed limit and the rate of change of distance to the preceding vehicle; and driver features, including the driver's head posture and line of sight. Big data processing sub-unit: Aggregates data through the cloud platform to build a driving behavior feature database, and uses data mining algorithms to generate driving behavior patterns. This includes using statistical analysis and clustering algorithms to generate regular behavior patterns, and using association rule algorithms and anomaly detection algorithms to generate abnormal behavior patterns. The data is then processed using Min-Max standardization and stored in a multidimensional feature table and rule condition table.

9. The device according to claim 8, characterized in that In the system architecture unit: The vehicle-side equipment includes an on-board processor and a storage device. The on-board processor is used to run real-time judgment and feedback programs, and the storage device is used to store temporary data and some key algorithms. The cloud platform includes large-scale data storage servers and high-performance computing clusters for storing massive amounts of driving behavior data and conducting complex data training and analysis; The edge computing unit includes an on-board edge computing chip and related software modules to realize real-time data preprocessing and preliminary judgment functions, and data is transmitted between the vehicle-side equipment, cloud platform and edge computing unit through a secure communication network.

Images