Executive Summary

Estimating the driver behavior using simulation tools is considered to be one of the most influential practices amongst many scientists and researchers.  Since the 19^th century, many researchers have successfully proposed methods and models to analyze the driving behavior through traffic simulations. These research and development studies include various models categorizes as integrated and independent, characteristic and behavior recognition/measurements methods, vehicle driving safety analyses tools and traffic regulatory proposals.

In this paper, a detailed literature review was conducted to analyze the research work completed/published in the last many decades. Through research, it was indicated that the recent developments in the driver behavior simulation have been productive due to the integrated model designing. Safety while driving has also been indicated to be one of the prime objectives of many researchers. Models developed to provide valuable data can be used to reduce the traffic risks and accidents.

Among the many research papers reviewed to prepare this paper, four research articles were analyzed to study the Driver Behavior in Traffic Simulation.

Introduction

Vehicle driving behavior model can be described as the presentation response of the vehicle being derived under various traffic conditions. Driving behavior models and simulations are developed to estimate the possible flow of the traffic through analyzing the flow characteristics of group of vehicles. The models use the variable such as speed, acceleration, lane changing, safety and other variable such as brake application and characteristic of aggregated traffic flow. The application of these models also includes the safety analyses and capacity estimates.

Much research has been conducted in the past and various driving models were developed. Models such as described by Rothery in 1997 presented the vehicle behavior while it is following a leader vehicle. More recently, models have been developed through microscopic traffic simulations which presented acceleration models, and with unconditional driver behavior with respect to following a leader vehicle. Accelerations models (Ahmed, 1999; Zhang et al., 1998; Yang, 1996) defined models which considered the driving behaviors with respect to lane changing, emergency, free flow and car following such as acceleration/deceleration and reactivations. The driving regimes considered in these papers were modeled independently, such as for driver/vehicle under acceleration focus on attaining the desired speed. 

Research conducted by Toledo et al, published in 2007 developed and implemented methods of integrated driving behavior modeling. This research presented the models which were integrated with short term goals and plans. It was assumed that the driver seek to complete the short term goals, such as the lane changing, through attaining the short term plan of achieving a target gap between the vehicles. The driver then facilitated the plan by acceleration and deceleration. The research captured such dependencies and estimated an integrated model.

Methods which could capture these variable and decisions were discussed and analyzed in detail.  The driver’s capability of planning plays a key role in application of these methods. The inter dependencies in many cases are conditional to the decision levels, for example if a driver plans to change the lane, only then the drive implies the acceleration or deceleration process. Therefore the specification of the lower level decision are integrated into the higher level decisions, which eventually captures the influence of inter dependencies.

Both the mandatory lane changing and discretionary lane changing behaviors were integrated and implemented in a single model/framework, which allows balancing lane changing considerations.  Similarly for accelerations, various possible driver behaviors were considered to develop an integrated model. Possibilities of acceleration with respect to attaining the target gap, changing the lane and/or keeping up with leading vehicle speed were considered. The framework adopted for acceleration was such that it could react differently for each case. Specifications such as the reaction time headway-threshold were implied in the model.

Furthermore, case studies completed in the paper showed that further developments are required for specific acceleration cases.  Following are the results obtained for tariff speed again time period for two locations UK network of Southampton.

Figure-1: Case study results for the two locations (Source: Toledo et al, 2007)

From the results, it can be seen that the traffic speeds are higher in integrated models, which is because of faster congestion built up rates with independent models in comparison with observed model and integrated model. Furthermore, it can be noticed that simulated traffic in independent models takes longer to recover the lost speed. The case studies showed that the integrated driving behavior model has the ability to perform better than the independent models developed for accelerations and lane changing.

Apart from the traffic flow, much consideration is given to the driver safety. Deery and Fildes (1999) presented research analyses on risk perception and young and beginner sub-types characteristics. Many young drivers along the globe are found to be involved in the traffic crashed, and the young driver safety has been discussed in many platforms in the recent past. Research conducted (Mayhew & Simpson, 1995) showed that the young driver’s performance are in many ways inferior as compared to the experienced drivers. The critical lack of skills in novice drivers are perception of hazard, lack of vehicle control with respect to response requirements and calibration.

Studies further show that the novice drivers are inclined to adopt more hazardous/risky driving styles. Data presented in many research articles (Evans and Wasielewski, 1983; Wasielewski, 1984) presented that the young drivers tend to leave shorter gap between the vehicles i.e. the gap between the subject vehicle and leading vehicle, and driving the vehicles at faster speeds. 

The research completed by Deery and Fildes (1999) included two related studies in which personality and driving related measures, and the driving performance of the young divers was analyzed through simulations. Using the five distinct sub-classes under analyses

Table-1; Results of five distinct sub-classes (Source: Hamish Deery and Fildes, 1999)

The results presented in the Table-1 show that for each sub-class, the measures of the variables are different. These results show the natural measured variations for all driver clusters.  The following figure-2 the results of average speed in km/h against the distance from the point of collision. Use of various average speeds and different range of distances from the point of collision were used to produce the results shown below

Figure-2: Results of five distinct sub-classes (Source: Deery and Fildes, 1999)

The results of the research indicated that the group of young/novice drivers cannot be considered as homogeneous.   The group of drivers was categorized into different sub-types, five in total, which was bases upon the personality level, the attitude towards driving and hostility/aggression measures. The research conducted showed that the group sub-types behavior differs with respect to attitude, style of driving, accident involvement/record, response to risks, and risk perception.  These results are considered to be persistent with various other research paper such as Jessor, 1987 and Wilson & Jonah, 1988.

Conclusion

Although the integrated driving behavior model can be considered to be advantageous in many applications, however these models are quite complex require intense computational input in designing the simulations.  Research conducted by various researchers showed that calibration required in modeling the driver’s behavior can be completed using computer simulations. This simulation can be completed using disaggregated data which defines the behavior model. Specific characteristics are integrated for the model parameters to complete the calibration process using the standardized aggregated data.  Safety of the drivers are dependent upon many variable include the experience of the divers. The applications such as ‘IVDRs’ can be used to obtain important information of driver’ behaviors with respect to their safety. Case studies analyses show that the safety is dependent upon various dependencies, and can be improved through applications of set of optimized regulations.

However, the results shown in various research papers suggest that further research and design is required to enhance the component model performances through optimized set of specifications.

References

Ahmed, K.I., 1999. Modeling drivers acceleration and lane changing behaviors, PhD thesis. Department of Civil and Environmental

Engineering, MIT.

Ahmed, K.I., Ben-Akiva, M., Koutsopoulos, H.N., and Mishalani, R.G., 1996. Models of freeway lane changing and gap acceptance

behavior. In: Proceedings of the 13th International Symposium on the Theory of Traffic Flow and Transportation, pp. 501–515.

Zhang, Y., Owen, L.E., Clark, J.E., 1998. A multi-regime approach for microscopic traffic simulation. Transportation Research Board, 77th Annual Meeting.

Yang, Q., Koutsopoulos, H.N., Ben-Akiva, M., 2000. Simulation laboratory for evaluating dynamic traffic management systems.

Transportation Research Record 1710, 122–130.

Toledo,T., Koutsopoulos, H.N. and Ben-Akiva, M., 2007. Integrated Driving Behavior Modeling. Transportation Research Part C 15.

Deery, H.A. and Fildes, B.N., 1999. Young Novice Driver Subtypes: Relationship to High-Risk Behavior, Traffic Accident. Record, and Simulator Driving Performance. Human Factors: The Journal of the Human Factors and Ergonomics Society 1999 41: 628.

Jensen, M., Alexander, K. and Wagner, J., 2011. Analysis of in-vehicle driver behaviour data for improved safety. Int. J. Vehicle Safety, Vol. 5, No. 3, 2011

Kuge, N., Yamamura, T., Shimoyama, O. and Liu, A., 1998.  A Driver Behavior Recognition Method Based on a Driver Model Framework. Society of Automotive Engineers, Inc. 2000-01-0349

Rothery, R.W., 1997. Car-following models. In: Gartner, N.H., Messer, C.J., Rathi, A.K. (Eds.), Monograph of Traffic Flow Theory.

Mayhew, D. R., & Simpson, H. M. (1995). The role of driving

experience: Implications for training and licensing of new drivers.

Toronto, Canada: Insurance Bureau of Canada

Evans, L., & Wasielewski, P. (1983). Risky driving related to driver

and vehicle characteristics. Accident Analysis and Prevention,

15, 121–136.

Wasielewski, P. (1984). Speed as a measure of driver risk: Observed

speed versus driver and vehicle characteristics. Accident

Analysis and Prevention, 16, 89–103