The basic theory of YOLO series object detection algorithms is discussed, the dangerous driving behavior dataset is collected and produced, and then the YOLOv7 network is introduced in detail, the deep separable convo...The basic theory of YOLO series object detection algorithms is discussed, the dangerous driving behavior dataset is collected and produced, and then the YOLOv7 network is introduced in detail, the deep separable convolution and CA attention mechanism are introduced, the YOLOv7 bounding box loss function and clustering algorithm are optimized, and the DB-YOLOv7 network structure is constructed. In the first stage of the experiment, the PASCAL VOC public dataset was utilized for pre-training. A comparative analysis was conducted to assess the recognition accuracy and inference time before and after the proposed improvements. The experimental results demonstrated an increase of 1.4% in the average recognition accuracy, alongside a reduction in the inference time by 4 ms. Subsequently, a model for the recognition of dangerous driving behaviors was trained using a specialized dangerous driving behavior dataset. A series of experiments were performed to evaluate the efficacy of the DB-YOLOv7 algorithm in this context. The findings indicate a significant enhancement in detection performance, with a 4% improvement in accuracy compared to the baseline network. Furthermore, the model’s inference time was reduced by 20%, from 25 ms to 20 ms. These results substantiate the effectiveness of the DB-YOLOv7 recognition algorithm for detecting dangerous driving behaviors, providing comprehensive validation of its practical applicability.展开更多
In order to give a new way for modeling driving behavior, identifying road traffic accident causation and solving a variety of road traffic safety problems such as driving errors prevention and driving behavior analys...In order to give a new way for modeling driving behavior, identifying road traffic accident causation and solving a variety of road traffic safety problems such as driving errors prevention and driving behavior analysis, a new driving behavior shaping model is proposed, which could be used to assess the degree of effect of driving error upon road traffic safety. Driver behavior shaping model based on driving reliability and safety analysis could be used to identify the road traffic accident causation, to supply data for driver's behavior training, to evaluate driving procedures, to human factor design of road traffic system.展开更多
Based on the driver surveillance video data and controller area network(CAN)data,the methods of studying commercial vehicles’driving behavior is relatively advanced.However,these methods have difficulty in covering p...Based on the driver surveillance video data and controller area network(CAN)data,the methods of studying commercial vehicles’driving behavior is relatively advanced.However,these methods have difficulty in covering private vehicles.Naturalistic driving studies have disadvantages of small sample size and high cost,one new driving behavior evaluation method using massive vehicle trajectory data is put forward.An automatic encoding machine is used to reduce the noise of raw data,and then driving dynamics and self-organizing mapping(SOM)classification are used to give thresholds or the judgement method of overspeed,rapid speed change,rapid turning and rapid lane changing.The proportion of different driving behaviors and typical dangerous driving behaviors is calculated,then the temporal and spatial distribution of drivers’driving behavior and the driving behavior characteristics on typical roads are analyzed.Driving behaviors on accident-prone road sections and normal road sections are compared.Results show that in Shenzhen,frequent lane changing and overspeed are the most common unsafe driving behaviors;16.1%drivers have relatively aggressive driving behavior;the proportion of dangerous driving behavior is higher outside the original economic special zone;dangerous driving behavior is highly correlated with traffic accident frequency.展开更多
To prevent and reduce road traffic accidents and improve driver safety awareness and bad driving be-haviors,we propose a safety evaluation method for commercial vehicle driving behavior.Three driving style clas-sifica...To prevent and reduce road traffic accidents and improve driver safety awareness and bad driving be-haviors,we propose a safety evaluation method for commercial vehicle driving behavior.Three driving style clas-sification indexes were extracted using driving data from commercial vehicles and four primary and ten secondary safety evaluation indicators.Based on the stability of commercial vehicles transporting goods,the acceleration index is divided into three levels according to the statistical third quartile,and the evaluation expression of the safety index evaluation is established.Drivers were divided into conservative,moderate,and radical using K-means++.The weights corresponding to each index were calculated using a combination of the analytic hierarchy process(AHP)and criteria importance through intercriteria correlation(CRITIC),and the driving behavior scores of various drivers were calculated according to the safety index score standard.The established AHP-CRITIC safety evaluation model was verified using the actual driving behavior data of commercial vehicle drivers.The calculation results show that the proposed evaluation model can clearly distinguish between the types of drivers with different driving styles,verifying its rationality and validity.The evaluation results can provide a reference for transportation management departments and enterprises.展开更多
The complexity of signal controlled traffic largely stems from the various driving behaviors developed in response to the traffic signal. However, the existing models take a few driving behaviors into account and cons...The complexity of signal controlled traffic largely stems from the various driving behaviors developed in response to the traffic signal. However, the existing models take a few driving behaviors into account and consequently the traffic dynamics has not been completely explored. Therefore, a new cellular automaton model, which incorporates the driving behaviors typically manifesting during the different stages when the vehicles are moving toward a traffic light, is proposed in this paper. Numerical simulations have demonstrated that the proposed model can produce the spontaneous traffic breakdown and the dissolution of the over-saturated traffic phenomena. Furthermore, the simulation results indicate that the slow-to-start behavior and the inch-forward behavior can foster the traffic breakdown. Particularly, it has been discovered that the over-saturated traffic can be revised to be an under-saturated state when the slow-down behavior is activated after the spontaneous breakdown. Finally, the contributions of the driving behaviors on the traffic breakdown have been examined.展开更多
In order to make full use of the driver’s long-term driving experience in the process of perception, interaction and vehicle control of road traffic information, a driving behavior rule extraction algorithm based on ...In order to make full use of the driver’s long-term driving experience in the process of perception, interaction and vehicle control of road traffic information, a driving behavior rule extraction algorithm based on artificial neural network interface(ANNI) and its integration is proposed. Firstly, based on the cognitive learning theory, the cognitive driving behavior model is established, and then the cognitive driving behavior is described and analyzed. Next, based on ANNI, the model and the rule extraction algorithm(ANNI-REA) are designed to explain not only the driving behavior but also the non-sequence. Rules have high fidelity and safety during driving without discretizing continuous input variables. The experimental results on the UCI standard data set and on the self-built driving behavior data set, show that the method is about 0.4% more accurate and about 10% less complex than the common C4.5-REA, Neuro-Rule and REFNE. Further, simulation experiments verify the correctness of the extracted driving rules and the effectiveness of the extraction based on cognitive driving behavior rules. In general, the several driving rules extracted fully reflect the execution mechanism of sequential activity of driving comprehensive cognition, which is of great significance for the traffic of mixed traffic flow under the network of vehicles and future research on unmanned driving.展开更多
In the light of the visual angle model(VAM),an improved car-following model considering driver's visual angle,anticipated time and stabilizing driving behavior is proposed so as to investigate how the driver's...In the light of the visual angle model(VAM),an improved car-following model considering driver's visual angle,anticipated time and stabilizing driving behavior is proposed so as to investigate how the driver's behavior factors affect the stability of the traffic flow.Based on the model,linear stability analysis is performed together with bifurcation analysis,whose corresponding stability condition is highly fit to the results of the linear analysis.Furthermore,the time-dependent Ginzburg–Landau(TDGL)equation and the modified Korteweg–de Vries(m Kd V)equation are derived by nonlinear analysis,and we obtain the relationship of the two equations through the comparison.Finally,parameter calibration and numerical simulation are conducted to verify the validity of the theoretical analysis,whose results are highly consistent with the theoretical analysis.展开更多
Abnormal driving behavior identification( ADBI) has become a research hotspot because of its significance in driver assistance systems. However,current methods still have some limitations in terms of accuracy and reli...Abnormal driving behavior identification( ADBI) has become a research hotspot because of its significance in driver assistance systems. However,current methods still have some limitations in terms of accuracy and reliability under severe traffic scenes. This paper proposes a new ADBI method based on direction and position offsets,where a two-factor identification strategy is proposed to improve the accuracy and reliability of ADBI. Self-adaptive edge detection based on Sobel operator is used to extract edge information of lanes. In order to enhance the efficiency and reliability of lane detection,an improved lane detection algorithm is proposed,where a Hough transform based on local search scope is employed to quickly detect the lane,and a validation scheme based on priori information is proposed to further verify the detected lane. Experimental results under various complex road conditions demonstrate the validity of the proposed ADBI.展开更多
Bus safety is a matter of great importance in many developing countries, with driving behaviors among bus drivers identified as a primary factor contributing to accidents. This concern is particularly amplified in mix...Bus safety is a matter of great importance in many developing countries, with driving behaviors among bus drivers identified as a primary factor contributing to accidents. This concern is particularly amplified in mixed traffic flow (MTF) environments with time pressure (TP). However, there is a lack of sufficient research exploring the relationships among these factors. This study consists of two papers that aim to investigate the impact of MTF environments with TP on the driving behaviors of bus drivers. While the first paper focuses on violated driving behaviors, this particular paper delves into mistake-prone driving behaviors (MDB). To collect data on MDB, as well as perceptions of MTF and TP, a questionnaire survey was implemented among bus drivers. Factor analyses were employed to create new measurements for validating MDB in MTF environments. The study utilized partial correlation and linear regression analyses with the Bayesian Model Averaging (BMA) method to explore the relationships between MDB and MTF/TP. The results revealed a modified scale for MDB. Two MTF factors and two TP factors were found to be significantly associated with MDB. A high presence of motorcycles and dangerous interactions among vehicles were not found to be associated with MDB among bus drivers. However, bus drivers who perceived motorcyclists as aggressive, considered road users’ traffic habits as unsafe, and perceived bus routes’ punctuality and organization as very strict were more likely to exhibit MDB. Moreover, the results from the three MDB predictive models demonstrated a positive impact of bus route organization on MDB among bus drivers. The study also examined various relationships between the socio-demographic characteristics of bus drivers and MDB. These findings are of practical significance in developing interventions aimed at reducing MDB among bus drivers operating in MTF environments with TP.展开更多
Electric vehicles are widely embraced as a promising solution to reduce energy consumption and emission to achieve the Carbon Peak and Carbon Neutrality vision,especially in developing countries.Specifically,it’s vit...Electric vehicles are widely embraced as a promising solution to reduce energy consumption and emission to achieve the Carbon Peak and Carbon Neutrality vision,especially in developing countries.Specifically,it’s vital important to understand the ecological performance of electric vehicles and its association with driving behaviors under varying road and environmental conditions.However,current researches on ecological driving behavior mostly use structured data to reflect the characteristics of ecological driving behavior,and it is difficult to accurately reveal the recessive relationship between driving behavior and energy consumption.One promising and prevalent method for comprehensively and in-depth characterizing driving behaviors is“graph spectrums”,which allows for an effective and illustrative representation of complex driving behavior characteristics.This study presented an assessment method of ecological driving for electric vehicles based on the graph.Firstly,a multi-source refined data set was constructed through naturalistic driving experiments(NDE).Four typical traffic state(CCCF:congested close car-following;CSSF:constrained slow free-flow;CSCF:constrained slow carfollowing;UFFF:unconstrained fast free-flow)were classified through longitudinal acceleration data,and driving behavior graph was constructed to realize the visual representation of driving behavior.Then,the energy consumption graph was constructed using the energy loss of 100 km(EL)index.After the six drivers with the highest and lowest ecological assessment of driving behavior using the behavior graph and energy consumption graph,proposing the quantitative analysis of fifteen drivers'ecology driving behavior.The results show that:1)The graphical method can describe the individual features of a driver’s ecological driving behavior;2)Rapid acceleration of driving behavior leads to high energy consumption;3)In the comparison among the six ecodrivers and energy-intensive drivers,founding that the energy-intensive drivers accelerate and decelerate significantly more in CCCF traffic state;4)The driving behavior was more complex and unecological in CCCF traffic state;5)Fifteen drivers had lower ecological scores in start-up driving.This study proposes a method for visualizing ecology driving behavior that not only help understand the individual characteristics of ecological driving behaviors,but also offers substantial application value for the subsequent construction of Ecological driving behavior regulation models.展开更多
Abnormal driving behavior includes driving distraction,fatigue,road anger,phone use,and an exceptionally happy mood.Detecting abnormal driving behavior in advance can avoid traffic accidents and reduce the risk of tra...Abnormal driving behavior includes driving distraction,fatigue,road anger,phone use,and an exceptionally happy mood.Detecting abnormal driving behavior in advance can avoid traffic accidents and reduce the risk of traffic conflicts.Traditional methods of detecting abnormal driving behavior include using wearable devices to monitor blood pressure,pulse,heart rate,blood oxygen,and other vital signs,and using eye trackers to monitor eye activity(such as eye closure,blinking frequency,etc.)to estimate whether the driver is excited,anxious,or distracted.Traditional monitoring methods can detect abnormal driving behavior to a certain extent,but they will affect the driver’s normal driving state,thereby introducing additional driving risks.This research uses the combined method of support vector machine and dlib algorithm to extract 68 facial feature points from the human face,and uses an SVM model as a strong classifier to classify different abnormal driving statuses.The combined method reaches high accuracy in detecting road anger and fatigue status and can be used in an intelligent vehicle cabin to improve the driving safety level.展开更多
Driver support and infotainment systems can be adapted to the specific needs of individual drivers by assessing driver skill and state.In this paper,we present a machine learning approach to classifying the skill at m...Driver support and infotainment systems can be adapted to the specific needs of individual drivers by assessing driver skill and state.In this paper,we present a machine learning approach to classifying the skill at maneuvering by drivers using both longitudinal and lateral controls in a vehicle.Conceptually,a model of drivers is constructed on the basis of sensor data related to the driving environment,the drivers'behaviors,and the vehi-cles'responses to the environment and behavior together.Once the model is built,the driving skills of an unknown driver can be classified automatically from the driving data.In this paper,we demonstrate the feasibility of using the proposed method to assess driving skill from the results of a driving simulator.We experiment with curve driving scenes,using both full curve and segmented curve sce-narios.Six curves with different radii and angular changes were set up for the experiment.In the full curve driving scene,principal component analysis and a support vector machine-based method accurately classified drivers in 95.7%of cases when using driving data about high-and low/average-skilled driver groups.In the cases with seg-mented curves,classification accuracy was 89%.展开更多
Collision avoidance decision-making models of multiple agents in virtual driving environment are studied. Based on the behavioral characteristics and hierarchical structure of the collision avoidance decision-making i...Collision avoidance decision-making models of multiple agents in virtual driving environment are studied. Based on the behavioral characteristics and hierarchical structure of the collision avoidance decision-making in real life driving, delphi approach and mathematical statistics method are introduced to construct pair-wise comparison judgment matrix of collision avoidance decision choices to each collision situation. Analytic hierarchy process （AHP） is adopted to establish the agents＇ collision avoidance decision-making model. To simulate drivers＇ characteristics, driver factors are added to categorize driving modes into impatient mode, normal mode, and the cautious mode. The results show that this model can simulate human＇s thinking process, and the agents in the virtual environment can deal with collision situations and make decisions to avoid collisions without intervention. The model can also reflect diversity and uncertainly of real life driving behaviors, and solves the multi-objective, multi-choice ranking priority problem in multi-vehicle collision scenarios. This collision avoidance model of multi-agents model is feasible and effective, and can provide richer and closer-to-life virtual scene for driving simulator, reflecting real-life traffic environment more truly, this model can also promote the practicality of driving simulator.展开更多
This paper is to explore the problems of intelligent connected vehicles(ICVs)autonomous driving decision-making under a 5G-V2X structured road environment.Through literature review and interviews with autonomous drivi...This paper is to explore the problems of intelligent connected vehicles(ICVs)autonomous driving decision-making under a 5G-V2X structured road environment.Through literature review and interviews with autonomous driving practitioners,this paper firstly puts forward a logical framework for designing a cerebrum-like autonomous driving system.Secondly,situated on this framework,it builds a hierarchical finite state machine(HFSM)model as well as a TOPSIS-GRA algorithm for making ICV autonomous driving decisions by employing a data fusion approach between the entropy weight method(EWM)and analytic hierarchy process method(AHP)and by employing a model fusion approach between the technique for order preference by similarity to an ideal solution(TOPSIS)and grey relational analysis(GRA).The HFSM model is composed of two layers:the global FSM model and the local FSM model.The decision of the former acts as partial input information of the latter and the result of the latter is sent forward to the local pathplanning module,meanwhile pulsating feedback to the former as real-time refresh data.To identify different traffic scenarios in a cerebrum-like way,the global FSM model is designed as 7 driving behavior states and 17 driving characteristic events,and the local FSM model is designed as 16 states and 8 characteristic events.In respect to designing a cerebrum-like algorithm for state transition,this paper firstly fuses AHP weight and EWM weight at their output layer to generate a synthetic weight coefficient for each characteristic event;then,it further fuses TOPSIS method and GRA method at the model building layer to obtain the implementable order of state transition.To verify the feasibility,reliability,and safety of theHFSMmodel aswell as its TOPSISGRA state transition algorithm,this paper elaborates on a series of simulative experiments conducted on the PreScan8.50 platform.The results display that the accuracy of obstacle detection gets 98%,lane line prediction is beyond 70 m,the speed of collision avoidance is higher than 45 km/h,the distance of collision avoidance is less than 5 m,path planning time for obstacle avoidance is averagely less than 50 ms,and brake deceleration is controlled under 6 m/s2.These technical indexes support that the driving states set and characteristic events set for the HFSM model as well as its TOPSIS-GRA algorithm may bring about cerebrum-like decision-making effectiveness for ICV autonomous driving under 5G-V2X intelligent road infrastructure.展开更多
Driver behavior modeling is becoming increasingly important in the study of traffic safety and devel- opment of cognitive vehicles. An algorithm for dealing with reliability for both digital driving and conventional d...Driver behavior modeling is becoming increasingly important in the study of traffic safety and devel- opment of cognitive vehicles. An algorithm for dealing with reliability for both digital driving and conventional driving has been developed in this paper. Problems of digital driving error classification, digital driving error probability quantification and digital driving reliability simulation have been addressed using a comparison re- search method. Simulation results show that driving reliability analysis discussed here is capable of identifying digital driving behavior characteristics and achieving safety assessment of intelligent transportation system.展开更多
The authors propose a two-stage method for recognizing driving situations on the basis of driving signals for application to a safe human interface of an in-vehicle information system. In first stage, an unknown drivi...The authors propose a two-stage method for recognizing driving situations on the basis of driving signals for application to a safe human interface of an in-vehicle information system. In first stage, an unknown driving situation is determined as stopping behavior or non-stopping behavior. In second stage, a Hidden Markov Model （HMM）-based pattern recognition method is used to model and recognize six non-stopping driving situations. The authors attempt to find the optimal HMM configuration to improve the performance of driving situation recognition. Center for Integrated Acoustic Information Research （CLAIR） in-vehicle corpus is used to evaluate the HMM-based recognition method. Driving situation categories are recognized using five driving signals. The proposed method achieves a relative error reduction rate of 30.9% compared to a conventional one-stage based HMMs.展开更多
Expressway tunnels are semi-enclosed structures characterized by monotonous alignment transitions and unique lighting environments,which can easily lead to drivers developing constrained and irritable psychology.This ...Expressway tunnels are semi-enclosed structures characterized by monotonous alignment transitions and unique lighting environments,which can easily lead to drivers developing constrained and irritable psychology.This may result in risky behaviors,e.g.,speeding and fatigued driving.Previous research on tunnel driving behaviors mainly focuses on visual factors,neglecting the impacts of nonstationary time-series combined parameters on risky driving.Firstly,30 drivers were recruited to carry out the real test.Then,based on the evolution of time series,drawing inspiration from the concept of lineage in biology,and considering multiple driving performance indicators,driving behavior chains and the feature spectrum were constructed.The characteristics of the behavior spectrum were divided into six groups:electroencephalogram,heart rate,eye movement,speed,steering,and carfollowing behaviors.Subsequently,the spectral analysis using the spectral radius property of matrix theory revealed the distinctive characteristics of risky driving behaviors.The study deeply explored the inducing mechanism,hidden patterns,and rules of risky driving behaviors under the coupling effect of tunnel environment and drivers’attributes.Finally,the significant features that influence driving behaviors were used as the input variables for constructing identification models using the adaptive boosting(AdaBoost)and random forest(RF)algorithms.The synthetic minority over-sampling technique(SMOTE)and adaptive synthetic sampling(ADASYN)were employed for oversampling.The results indicate that the ADASYN-RF algorithm outperformed others,achieving a precise recall rate area under the curve(AUPRC)of 0.978 when using the spectral radius of the speed and steering groups as input variables.These findings offer theoretical guidance for developing tunnel traffic safety strategies.展开更多
The aim of this paper is to understand the factors that influence unsafe driving practices by examining published studies that utilized the theory of planned behavior(TPB)to predict driving behavior.To this end,42 stu...The aim of this paper is to understand the factors that influence unsafe driving practices by examining published studies that utilized the theory of planned behavior(TPB)to predict driving behavior.To this end,42 studies published up to the end of 2021 are reviewed to evaluate the predictive utility of TPB by employing a meta-analysis and structural equation model.The results indicate that these studies sought to predict 20 distinct driving behaviors(e.g.,drink-driving,use of cellphone while driving,aggressive driving)using the original TPB constructs and 43 additional variables.The TPB model with the three original constructs is found to account for 32%intentional variance and 34%behavioral variance.Among the 43 variables researchers have examined in TPB studies related to driving behavior,this study identified the six that are commonly used to enhance the TPB model’s predictive power.These variables are past behavior,self-identity,descriptive norm,anticipated regret,risk perception,and moral norm.When past behavior is added to the original TPB model,it increases the explained variance in intention to 52%.When all six factors are added to the original TPB model,the best model has only four variables(perceived risk,self-identity,descriptive norm,and moral norm);and increases the explained variance to 48%.The influence of the TPB constructs on intention is modified by behavior category and traffic category.The findings of this paper validate the application of TPB to predicting driving behavior.It is the first study to do this through the use of meta-analysis and structural equation modeling.展开更多
To explore the relationship between rear-end crash risk and its influencing factors, on-road experiments were conducted for measuring the individual vehicle trajectory data associated with novice and experienced drive...To explore the relationship between rear-end crash risk and its influencing factors, on-road experiments were conducted for measuring the individual vehicle trajectory data associated with novice and experienced drivers. The rear-end crash potential probability based on the time to collision was proposed to represent the interpretation of rear-end crash risk.One-way analysis of variance was applied to compare the rearend crash risks for novice and experienced drivers. The rearend crash risk models for novice and experienced drivers were respectively developed to identify the effects of contributing factors on the driver rear-end crash risk. Also, the cumulative residual method was used to examine the goodness-of-fit of models. The results show that there is a significant difference in rear-end risk between the novice and experienced drivers.For the novice drivers, three risk factors including the traffic volume, the number of lanes and gender are found to significantly impact on the rear-end crash risk, while significant impact factors for experienced drivers are the vehicle speed and traffic volume. The rear-end crash risk models perform well based on the existing limited data samples.展开更多
A new emergency evacuation car-following model （EECM） is proposed. The model aims to capture the main characteristics of traffic flow and driver behavior under an emergency evacuation, and it is developed on the bas...A new emergency evacuation car-following model （EECM） is proposed. The model aims to capture the main characteristics of traffic flow and driver behavior under an emergency evacuation, and it is developed on the basis of minimum safety distances with parts of the drivers＇ abnormal behavior in a panic emergency situation. A thorough questionnaire survey is undertaken among drivers of different ages. Based on the results from the survey, a safety-distance car-following model is formulated by taking into account two new parameters： a differential distributing coefficient and a driver＇ s experiential decision coefficient, which are used to reflect variations of driving behaviors under an emergency evacuation situation when compared with regular conditions. The formulation and derivation of the new model, as well as its properties and applicability are discussed. A case study is presented to compare the car-following trajectories using observed data under regular peak-hour traffic conditions and theoretical EECM results. The results indicate the consistency of the analysis of assumptions on the EECM and observations.展开更多
文摘The basic theory of YOLO series object detection algorithms is discussed, the dangerous driving behavior dataset is collected and produced, and then the YOLOv7 network is introduced in detail, the deep separable convolution and CA attention mechanism are introduced, the YOLOv7 bounding box loss function and clustering algorithm are optimized, and the DB-YOLOv7 network structure is constructed. In the first stage of the experiment, the PASCAL VOC public dataset was utilized for pre-training. A comparative analysis was conducted to assess the recognition accuracy and inference time before and after the proposed improvements. The experimental results demonstrated an increase of 1.4% in the average recognition accuracy, alongside a reduction in the inference time by 4 ms. Subsequently, a model for the recognition of dangerous driving behaviors was trained using a specialized dangerous driving behavior dataset. A series of experiments were performed to evaluate the efficacy of the DB-YOLOv7 algorithm in this context. The findings indicate a significant enhancement in detection performance, with a 4% improvement in accuracy compared to the baseline network. Furthermore, the model’s inference time was reduced by 20%, from 25 ms to 20 ms. These results substantiate the effectiveness of the DB-YOLOv7 recognition algorithm for detecting dangerous driving behaviors, providing comprehensive validation of its practical applicability.
文摘In order to give a new way for modeling driving behavior, identifying road traffic accident causation and solving a variety of road traffic safety problems such as driving errors prevention and driving behavior analysis, a new driving behavior shaping model is proposed, which could be used to assess the degree of effect of driving error upon road traffic safety. Driver behavior shaping model based on driving reliability and safety analysis could be used to identify the road traffic accident causation, to supply data for driver's behavior training, to evaluate driving procedures, to human factor design of road traffic system.
基金The National Natural Science Foundation of China(No.71641005)the National Key Research and Development Program of China(No.2018YFB1601105)
文摘Based on the driver surveillance video data and controller area network(CAN)data,the methods of studying commercial vehicles’driving behavior is relatively advanced.However,these methods have difficulty in covering private vehicles.Naturalistic driving studies have disadvantages of small sample size and high cost,one new driving behavior evaluation method using massive vehicle trajectory data is put forward.An automatic encoding machine is used to reduce the noise of raw data,and then driving dynamics and self-organizing mapping(SOM)classification are used to give thresholds or the judgement method of overspeed,rapid speed change,rapid turning and rapid lane changing.The proportion of different driving behaviors and typical dangerous driving behaviors is calculated,then the temporal and spatial distribution of drivers’driving behavior and the driving behavior characteristics on typical roads are analyzed.Driving behaviors on accident-prone road sections and normal road sections are compared.Results show that in Shenzhen,frequent lane changing and overspeed are the most common unsafe driving behaviors;16.1%drivers have relatively aggressive driving behavior;the proportion of dangerous driving behavior is higher outside the original economic special zone;dangerous driving behavior is highly correlated with traffic accident frequency.
基金the Natural Science Foundation of Guangxi(No.2020GXNSFDA238011)the Open Fund Project of Guangxi Key Laboratory of Automation Detection Technology and Instrument(No.YQ21203)the Independent Research Project of Guangxi Key Laboratory of Auto Parts and Vehicle Technology(No.2020GKLACVTZZ02)。
文摘To prevent and reduce road traffic accidents and improve driver safety awareness and bad driving be-haviors,we propose a safety evaluation method for commercial vehicle driving behavior.Three driving style clas-sification indexes were extracted using driving data from commercial vehicles and four primary and ten secondary safety evaluation indicators.Based on the stability of commercial vehicles transporting goods,the acceleration index is divided into three levels according to the statistical third quartile,and the evaluation expression of the safety index evaluation is established.Drivers were divided into conservative,moderate,and radical using K-means++.The weights corresponding to each index were calculated using a combination of the analytic hierarchy process(AHP)and criteria importance through intercriteria correlation(CRITIC),and the driving behavior scores of various drivers were calculated according to the safety index score standard.The established AHP-CRITIC safety evaluation model was verified using the actual driving behavior data of commercial vehicle drivers.The calculation results show that the proposed evaluation model can clearly distinguish between the types of drivers with different driving styles,verifying its rationality and validity.The evaluation results can provide a reference for transportation management departments and enterprises.
基金supported by the National Basic Research Program of China(Grand No.2012CB723303)the Beijing Committee of Science and Technology,China(Grand No.Z1211000003120100)
文摘The complexity of signal controlled traffic largely stems from the various driving behaviors developed in response to the traffic signal. However, the existing models take a few driving behaviors into account and consequently the traffic dynamics has not been completely explored. Therefore, a new cellular automaton model, which incorporates the driving behaviors typically manifesting during the different stages when the vehicles are moving toward a traffic light, is proposed in this paper. Numerical simulations have demonstrated that the proposed model can produce the spontaneous traffic breakdown and the dissolution of the over-saturated traffic phenomena. Furthermore, the simulation results indicate that the slow-to-start behavior and the inch-forward behavior can foster the traffic breakdown. Particularly, it has been discovered that the over-saturated traffic can be revised to be an under-saturated state when the slow-down behavior is activated after the spontaneous breakdown. Finally, the contributions of the driving behaviors on the traffic breakdown have been examined.
基金Project(2017YFB0102503)supported by the National Key Research and Development Program of ChinaProjects(U1664258,51875255,61601203)supported by the National Natural Science Foundation of China+1 种基金Projects(DZXX-048,2018-TD-GDZB-022)supported by the Jiangsu Province’s Six Talent Peak,ChinaProject(18KJA580002)supported by Major Natural Science Research Project of Higher Learning in Jiangsu Province,China
文摘In order to make full use of the driver’s long-term driving experience in the process of perception, interaction and vehicle control of road traffic information, a driving behavior rule extraction algorithm based on artificial neural network interface(ANNI) and its integration is proposed. Firstly, based on the cognitive learning theory, the cognitive driving behavior model is established, and then the cognitive driving behavior is described and analyzed. Next, based on ANNI, the model and the rule extraction algorithm(ANNI-REA) are designed to explain not only the driving behavior but also the non-sequence. Rules have high fidelity and safety during driving without discretizing continuous input variables. The experimental results on the UCI standard data set and on the self-built driving behavior data set, show that the method is about 0.4% more accurate and about 10% less complex than the common C4.5-REA, Neuro-Rule and REFNE. Further, simulation experiments verify the correctness of the extracted driving rules and the effectiveness of the extraction based on cognitive driving behavior rules. In general, the several driving rules extracted fully reflect the execution mechanism of sequential activity of driving comprehensive cognition, which is of great significance for the traffic of mixed traffic flow under the network of vehicles and future research on unmanned driving.
基金the Natural Science Foundation of Zhejiang Province,China(Grant Nos.LY22G010001,LY20G010004)the Program of Humanities and Social Science of Education Ministry of China(Grant No.20YJA630008)+1 种基金the National Key Research and Development Program of China-Traffic Modeling,Surveillance and Control with Connected&Automated Vehicles(Grant No.2017YFE9134700)the K.C.Wong Magna Fund in Ningbo University,China。
文摘In the light of the visual angle model(VAM),an improved car-following model considering driver's visual angle,anticipated time and stabilizing driving behavior is proposed so as to investigate how the driver's behavior factors affect the stability of the traffic flow.Based on the model,linear stability analysis is performed together with bifurcation analysis,whose corresponding stability condition is highly fit to the results of the linear analysis.Furthermore,the time-dependent Ginzburg–Landau(TDGL)equation and the modified Korteweg–de Vries(m Kd V)equation are derived by nonlinear analysis,and we obtain the relationship of the two equations through the comparison.Finally,parameter calibration and numerical simulation are conducted to verify the validity of the theoretical analysis,whose results are highly consistent with the theoretical analysis.
基金Supported by the National Natural Science Foundation of China(No.61304205,61502240)Natural Science Foundation of Jiangsu Province(BK20141002)+1 种基金Innovation and Entrepreneurship Training Project of College Students(No.201710300051,201710300050)Foundation for Excellent Undergraduate Dissertation(Design) of Naning University of Information Science & Technology
文摘Abnormal driving behavior identification( ADBI) has become a research hotspot because of its significance in driver assistance systems. However,current methods still have some limitations in terms of accuracy and reliability under severe traffic scenes. This paper proposes a new ADBI method based on direction and position offsets,where a two-factor identification strategy is proposed to improve the accuracy and reliability of ADBI. Self-adaptive edge detection based on Sobel operator is used to extract edge information of lanes. In order to enhance the efficiency and reliability of lane detection,an improved lane detection algorithm is proposed,where a Hough transform based on local search scope is employed to quickly detect the lane,and a validation scheme based on priori information is proposed to further verify the detected lane. Experimental results under various complex road conditions demonstrate the validity of the proposed ADBI.
文摘Bus safety is a matter of great importance in many developing countries, with driving behaviors among bus drivers identified as a primary factor contributing to accidents. This concern is particularly amplified in mixed traffic flow (MTF) environments with time pressure (TP). However, there is a lack of sufficient research exploring the relationships among these factors. This study consists of two papers that aim to investigate the impact of MTF environments with TP on the driving behaviors of bus drivers. While the first paper focuses on violated driving behaviors, this particular paper delves into mistake-prone driving behaviors (MDB). To collect data on MDB, as well as perceptions of MTF and TP, a questionnaire survey was implemented among bus drivers. Factor analyses were employed to create new measurements for validating MDB in MTF environments. The study utilized partial correlation and linear regression analyses with the Bayesian Model Averaging (BMA) method to explore the relationships between MDB and MTF/TP. The results revealed a modified scale for MDB. Two MTF factors and two TP factors were found to be significantly associated with MDB. A high presence of motorcycles and dangerous interactions among vehicles were not found to be associated with MDB among bus drivers. However, bus drivers who perceived motorcyclists as aggressive, considered road users’ traffic habits as unsafe, and perceived bus routes’ punctuality and organization as very strict were more likely to exhibit MDB. Moreover, the results from the three MDB predictive models demonstrated a positive impact of bus route organization on MDB among bus drivers. The study also examined various relationships between the socio-demographic characteristics of bus drivers and MDB. These findings are of practical significance in developing interventions aimed at reducing MDB among bus drivers operating in MTF environments with TP.
基金supported by the National Key R&D Program of China(2023YFC3081700)the National Natural Science Foundation of China(52372341).
文摘Electric vehicles are widely embraced as a promising solution to reduce energy consumption and emission to achieve the Carbon Peak and Carbon Neutrality vision,especially in developing countries.Specifically,it’s vital important to understand the ecological performance of electric vehicles and its association with driving behaviors under varying road and environmental conditions.However,current researches on ecological driving behavior mostly use structured data to reflect the characteristics of ecological driving behavior,and it is difficult to accurately reveal the recessive relationship between driving behavior and energy consumption.One promising and prevalent method for comprehensively and in-depth characterizing driving behaviors is“graph spectrums”,which allows for an effective and illustrative representation of complex driving behavior characteristics.This study presented an assessment method of ecological driving for electric vehicles based on the graph.Firstly,a multi-source refined data set was constructed through naturalistic driving experiments(NDE).Four typical traffic state(CCCF:congested close car-following;CSSF:constrained slow free-flow;CSCF:constrained slow carfollowing;UFFF:unconstrained fast free-flow)were classified through longitudinal acceleration data,and driving behavior graph was constructed to realize the visual representation of driving behavior.Then,the energy consumption graph was constructed using the energy loss of 100 km(EL)index.After the six drivers with the highest and lowest ecological assessment of driving behavior using the behavior graph and energy consumption graph,proposing the quantitative analysis of fifteen drivers'ecology driving behavior.The results show that:1)The graphical method can describe the individual features of a driver’s ecological driving behavior;2)Rapid acceleration of driving behavior leads to high energy consumption;3)In the comparison among the six ecodrivers and energy-intensive drivers,founding that the energy-intensive drivers accelerate and decelerate significantly more in CCCF traffic state;4)The driving behavior was more complex and unecological in CCCF traffic state;5)Fifteen drivers had lower ecological scores in start-up driving.This study proposes a method for visualizing ecology driving behavior that not only help understand the individual characteristics of ecological driving behaviors,but also offers substantial application value for the subsequent construction of Ecological driving behavior regulation models.
文摘Abnormal driving behavior includes driving distraction,fatigue,road anger,phone use,and an exceptionally happy mood.Detecting abnormal driving behavior in advance can avoid traffic accidents and reduce the risk of traffic conflicts.Traditional methods of detecting abnormal driving behavior include using wearable devices to monitor blood pressure,pulse,heart rate,blood oxygen,and other vital signs,and using eye trackers to monitor eye activity(such as eye closure,blinking frequency,etc.)to estimate whether the driver is excited,anxious,or distracted.Traditional monitoring methods can detect abnormal driving behavior to a certain extent,but they will affect the driver’s normal driving state,thereby introducing additional driving risks.This research uses the combined method of support vector machine and dlib algorithm to extract 68 facial feature points from the human face,and uses an SVM model as a strong classifier to classify different abnormal driving statuses.The combined method reaches high accuracy in detecting road anger and fatigue status and can be used in an intelligent vehicle cabin to improve the driving safety level.
文摘Driver support and infotainment systems can be adapted to the specific needs of individual drivers by assessing driver skill and state.In this paper,we present a machine learning approach to classifying the skill at maneuvering by drivers using both longitudinal and lateral controls in a vehicle.Conceptually,a model of drivers is constructed on the basis of sensor data related to the driving environment,the drivers'behaviors,and the vehi-cles'responses to the environment and behavior together.Once the model is built,the driving skills of an unknown driver can be classified automatically from the driving data.In this paper,we demonstrate the feasibility of using the proposed method to assess driving skill from the results of a driving simulator.We experiment with curve driving scenes,using both full curve and segmented curve sce-narios.Six curves with different radii and angular changes were set up for the experiment.In the full curve driving scene,principal component analysis and a support vector machine-based method accurately classified drivers in 95.7%of cases when using driving data about high-and low/average-skilled driver groups.In the cases with seg-mented curves,classification accuracy was 89%.
基金supported by National Basic Research Program (973 Program,No.2004CB719402)National Natural Science Foundation of China (No.60736019)Natural Science Foundation of Zhejiang Province, China(No.Y105430).
文摘Collision avoidance decision-making models of multiple agents in virtual driving environment are studied. Based on the behavioral characteristics and hierarchical structure of the collision avoidance decision-making in real life driving, delphi approach and mathematical statistics method are introduced to construct pair-wise comparison judgment matrix of collision avoidance decision choices to each collision situation. Analytic hierarchy process （AHP） is adopted to establish the agents＇ collision avoidance decision-making model. To simulate drivers＇ characteristics, driver factors are added to categorize driving modes into impatient mode, normal mode, and the cautious mode. The results show that this model can simulate human＇s thinking process, and the agents in the virtual environment can deal with collision situations and make decisions to avoid collisions without intervention. The model can also reflect diversity and uncertainly of real life driving behaviors, and solves the multi-objective, multi-choice ranking priority problem in multi-vehicle collision scenarios. This collision avoidance model of multi-agents model is feasible and effective, and can provide richer and closer-to-life virtual scene for driving simulator, reflecting real-life traffic environment more truly, this model can also promote the practicality of driving simulator.
基金funded by Chongqing Science and Technology Bureau (No.cstc2021jsyj-yzysbAX0008)Chongqing University of Arts and Sciences (No.P2021JG13)2021 Humanities and Social Sciences Program of Chongqing Education Commission (No.21SKGH227).
文摘This paper is to explore the problems of intelligent connected vehicles(ICVs)autonomous driving decision-making under a 5G-V2X structured road environment.Through literature review and interviews with autonomous driving practitioners,this paper firstly puts forward a logical framework for designing a cerebrum-like autonomous driving system.Secondly,situated on this framework,it builds a hierarchical finite state machine(HFSM)model as well as a TOPSIS-GRA algorithm for making ICV autonomous driving decisions by employing a data fusion approach between the entropy weight method(EWM)and analytic hierarchy process method(AHP)and by employing a model fusion approach between the technique for order preference by similarity to an ideal solution(TOPSIS)and grey relational analysis(GRA).The HFSM model is composed of two layers:the global FSM model and the local FSM model.The decision of the former acts as partial input information of the latter and the result of the latter is sent forward to the local pathplanning module,meanwhile pulsating feedback to the former as real-time refresh data.To identify different traffic scenarios in a cerebrum-like way,the global FSM model is designed as 7 driving behavior states and 17 driving characteristic events,and the local FSM model is designed as 16 states and 8 characteristic events.In respect to designing a cerebrum-like algorithm for state transition,this paper firstly fuses AHP weight and EWM weight at their output layer to generate a synthetic weight coefficient for each characteristic event;then,it further fuses TOPSIS method and GRA method at the model building layer to obtain the implementable order of state transition.To verify the feasibility,reliability,and safety of theHFSMmodel aswell as its TOPSISGRA state transition algorithm,this paper elaborates on a series of simulative experiments conducted on the PreScan8.50 platform.The results display that the accuracy of obstacle detection gets 98%,lane line prediction is beyond 70 m,the speed of collision avoidance is higher than 45 km/h,the distance of collision avoidance is less than 5 m,path planning time for obstacle avoidance is averagely less than 50 ms,and brake deceleration is controlled under 6 m/s2.These technical indexes support that the driving states set and characteristic events set for the HFSM model as well as its TOPSIS-GRA algorithm may bring about cerebrum-like decision-making effectiveness for ICV autonomous driving under 5G-V2X intelligent road infrastructure.
基金Sponsored by the National Natural Science Foundation of China(50878023)the Scientific Research Foundation for the Returned Overseas Chinese Scholars
文摘Driver behavior modeling is becoming increasingly important in the study of traffic safety and devel- opment of cognitive vehicles. An algorithm for dealing with reliability for both digital driving and conventional driving has been developed in this paper. Problems of digital driving error classification, digital driving error probability quantification and digital driving reliability simulation have been addressed using a comparison re- search method. Simulation results show that driving reliability analysis discussed here is capable of identifying digital driving behavior characteristics and achieving safety assessment of intelligent transportation system.
文摘The authors propose a two-stage method for recognizing driving situations on the basis of driving signals for application to a safe human interface of an in-vehicle information system. In first stage, an unknown driving situation is determined as stopping behavior or non-stopping behavior. In second stage, a Hidden Markov Model （HMM）-based pattern recognition method is used to model and recognize six non-stopping driving situations. The authors attempt to find the optimal HMM configuration to improve the performance of driving situation recognition. Center for Integrated Acoustic Information Research （CLAIR） in-vehicle corpus is used to evaluate the HMM-based recognition method. Driving situation categories are recognized using five driving signals. The proposed method achieves a relative error reduction rate of 30.9% compared to a conventional one-stage based HMMs.
基金supported by the National Natural Science Foundation of China(No.51978069)Science and Technology Project of Shandong Transportation Department(No.2022-KJ-044)+1 种基金Key Research and Development Plan of Shaanxi Province(No.2021KWZ-09)the Fundamental Research Funds for the Centrl Universities,CHD(No.300102342202).
文摘Expressway tunnels are semi-enclosed structures characterized by monotonous alignment transitions and unique lighting environments,which can easily lead to drivers developing constrained and irritable psychology.This may result in risky behaviors,e.g.,speeding and fatigued driving.Previous research on tunnel driving behaviors mainly focuses on visual factors,neglecting the impacts of nonstationary time-series combined parameters on risky driving.Firstly,30 drivers were recruited to carry out the real test.Then,based on the evolution of time series,drawing inspiration from the concept of lineage in biology,and considering multiple driving performance indicators,driving behavior chains and the feature spectrum were constructed.The characteristics of the behavior spectrum were divided into six groups:electroencephalogram,heart rate,eye movement,speed,steering,and carfollowing behaviors.Subsequently,the spectral analysis using the spectral radius property of matrix theory revealed the distinctive characteristics of risky driving behaviors.The study deeply explored the inducing mechanism,hidden patterns,and rules of risky driving behaviors under the coupling effect of tunnel environment and drivers’attributes.Finally,the significant features that influence driving behaviors were used as the input variables for constructing identification models using the adaptive boosting(AdaBoost)and random forest(RF)algorithms.The synthetic minority over-sampling technique(SMOTE)and adaptive synthetic sampling(ADASYN)were employed for oversampling.The results indicate that the ADASYN-RF algorithm outperformed others,achieving a precise recall rate area under the curve(AUPRC)of 0.978 when using the spectral radius of the speed and steering groups as input variables.These findings offer theoretical guidance for developing tunnel traffic safety strategies.
文摘The aim of this paper is to understand the factors that influence unsafe driving practices by examining published studies that utilized the theory of planned behavior(TPB)to predict driving behavior.To this end,42 studies published up to the end of 2021 are reviewed to evaluate the predictive utility of TPB by employing a meta-analysis and structural equation model.The results indicate that these studies sought to predict 20 distinct driving behaviors(e.g.,drink-driving,use of cellphone while driving,aggressive driving)using the original TPB constructs and 43 additional variables.The TPB model with the three original constructs is found to account for 32%intentional variance and 34%behavioral variance.Among the 43 variables researchers have examined in TPB studies related to driving behavior,this study identified the six that are commonly used to enhance the TPB model’s predictive power.These variables are past behavior,self-identity,descriptive norm,anticipated regret,risk perception,and moral norm.When past behavior is added to the original TPB model,it increases the explained variance in intention to 52%.When all six factors are added to the original TPB model,the best model has only four variables(perceived risk,self-identity,descriptive norm,and moral norm);and increases the explained variance to 48%.The influence of the TPB constructs on intention is modified by behavior category and traffic category.The findings of this paper validate the application of TPB to predicting driving behavior.It is the first study to do this through the use of meta-analysis and structural equation modeling.
基金The National Natural Science Foundation of China(No.51478110)
文摘To explore the relationship between rear-end crash risk and its influencing factors, on-road experiments were conducted for measuring the individual vehicle trajectory data associated with novice and experienced drivers. The rear-end crash potential probability based on the time to collision was proposed to represent the interpretation of rear-end crash risk.One-way analysis of variance was applied to compare the rearend crash risks for novice and experienced drivers. The rearend crash risk models for novice and experienced drivers were respectively developed to identify the effects of contributing factors on the driver rear-end crash risk. Also, the cumulative residual method was used to examine the goodness-of-fit of models. The results show that there is a significant difference in rear-end risk between the novice and experienced drivers.For the novice drivers, three risk factors including the traffic volume, the number of lanes and gender are found to significantly impact on the rear-end crash risk, while significant impact factors for experienced drivers are the vehicle speed and traffic volume. The rear-end crash risk models perform well based on the existing limited data samples.
基金The National Key Technology R&D Program of China during the 10th Five-Year Plan Period(No.2005BA41B11)the National Natural Science Foundation of China(No.50578003)
文摘A new emergency evacuation car-following model （EECM） is proposed. The model aims to capture the main characteristics of traffic flow and driver behavior under an emergency evacuation, and it is developed on the basis of minimum safety distances with parts of the drivers＇ abnormal behavior in a panic emergency situation. A thorough questionnaire survey is undertaken among drivers of different ages. Based on the results from the survey, a safety-distance car-following model is formulated by taking into account two new parameters： a differential distributing coefficient and a driver＇ s experiential decision coefficient, which are used to reflect variations of driving behaviors under an emergency evacuation situation when compared with regular conditions. The formulation and derivation of the new model, as well as its properties and applicability are discussed. A case study is presented to compare the car-following trajectories using observed data under regular peak-hour traffic conditions and theoretical EECM results. The results indicate the consistency of the analysis of assumptions on the EECM and observations.
