Professional Documents
Culture Documents
3, MARCH 2018
Abstract— Driver driving style plays an important role in vehi- safe and adequate performance, but also to adjust to the
cle energy management as well as driving safety. Furthermore, drivers’ needs, potentiate their acceptability and ultimately
it is key for advance driver assistance systems development, meet drivers’ preferences in a safe environment. Therefore,
toward increasing levels of vehicle automation. This fact has moti-
vated numerous research and development efforts on driving style it is essential the recognition of driving style (DS) and driver
identification and classification. This paper provides a survey on intention inference for the integration and development of
driving style characterization and recognition revising a variety these systems [1]–[7]. This statement is supported by the
of algorithms, with particular emphasis on machine learning recently proposed parallel driving framework based on cyber-
approaches based on current and future trends. Applications physical-social systems, where driver behavior and cognition
of driving style recognition to intelligent vehicle controls are
also briefly discussed, including experts’ predictions of the future was one of the key elements [8]–[10]. Furthermore, the society
development. is becoming more aware of environmental concerns, which
influences decisions made by industry and end users [11], [12].
Index Terms— Driving style, driving conditions, driver
behavior, driving style recognition, machine learning, intelligent Particularly, in 2004 27% of the global CO2 emissions were
vehicle control, energy efficiency, driving safety. caused by the transport sector, being 83% responsibility of
road vehicles [13]. Eco-friendliness, fuel supply security and
N OMENCLATURE CO2 regulations are driving research towards more efficient
powertrains including: conventional, battery electrical vehi-
ADAS Advanced Driver Assisted Systems
cles (BEVs), hybrid electric vehicles (HEVs), plug-in hybrid
BEV Battery Electric Vehicle
electric vehicles (PHEVs) between others [12].
DS Driving Style
Besides, industry has taken a further step to influence
FL Fuzzy Logic
driver driving style through active and passive corrective
GIS Geographical Information System
feedback [14] towards safer and eco-friendly practises. The
GMM Gaussian Mixture Model
firsts can intervene into the driving task directly by generating
GPS Global Positioning System
haptic inputs [2], [3], [7], [9], whilst the second is only
HEV Hybrid Electric Vehicle
advisory and targets to improve drivers’ awareness through
kNN k-nearest neighbor
visual or audio advice [15], [16]. These systems that provide
NN Neural Network
only feedback to the driver, either haptic or passive, are clas-
PHEV Plug-in Hybrid Electric Vehicle
sified as passive, whilst active systems can interfere directly
RB Rule-based
in the vehicle operation by modifying driver’s power demand
SVM Support Vector Machine
and recalibrating the components response to compensate for
deficient driving habits [13], [16]. Although the feedback suit-
I. I NTRODUCTION ability and effectiveness can be argued, it has been observed
Authorized licensed use limited to: Universitaetsbibliothek der RWTH Aachen. Downloaded on May 02,2021 at 19:16:50 UTC from IEEE Xplore. Restrictions apply.
MARINA MARTINEZ et al.: DRIVING STYLE RECOGNITION FOR INTELLIGENT VEHICLE CONTROL AND ADVANCED DRIVER ASSISTANCE 667
Authorized licensed use limited to: Universitaetsbibliothek der RWTH Aachen. Downloaded on May 02,2021 at 19:16:50 UTC from IEEE Xplore. Restrictions apply.
668 IEEE TRANSACTIONS ON INTELLIGENT TRANSPORTATION SYSTEMS, VOL. 19, NO. 3, MARCH 2018
Authorized licensed use limited to: Universitaetsbibliothek der RWTH Aachen. Downloaded on May 02,2021 at 19:16:50 UTC from IEEE Xplore. Restrictions apply.
MARINA MARTINEZ et al.: DRIVING STYLE RECOGNITION FOR INTELLIGENT VEHICLE CONTROL AND ADVANCED DRIVER ASSISTANCE 669
Authorized licensed use limited to: Universitaetsbibliothek der RWTH Aachen. Downloaded on May 02,2021 at 19:16:50 UTC from IEEE Xplore. Restrictions apply.
670 IEEE TRANSACTIONS ON INTELLIGENT TRANSPORTATION SYSTEMS, VOL. 19, NO. 3, MARCH 2018
Authorized licensed use limited to: Universitaetsbibliothek der RWTH Aachen. Downloaded on May 02,2021 at 19:16:50 UTC from IEEE Xplore. Restrictions apply.
MARINA MARTINEZ et al.: DRIVING STYLE RECOGNITION FOR INTELLIGENT VEHICLE CONTROL AND ADVANCED DRIVER ASSISTANCE 671
Authorized licensed use limited to: Universitaetsbibliothek der RWTH Aachen. Downloaded on May 02,2021 at 19:16:50 UTC from IEEE Xplore. Restrictions apply.
672 IEEE TRANSACTIONS ON INTELLIGENT TRANSPORTATION SYSTEMS, VOL. 19, NO. 3, MARCH 2018
their comparison with real drivers, which again necessitates is k-nearest neighbour (kNN), which is based on similarity
extensive data collection [32]. measures rated by the “votes” of the predefined k-neighbours.
This method was exploited by Vaitkus et al. [43] using a
C. Implemented Through Learning Algorithms total of five features extracted from the data, which provided
Threshold definition in RB algorithms conditions the results a 100% classification success using experimental data from
robustness and requires extensive data analysis. Data-driven 110 routes. Furthermore, the authors claimed 98% classifica-
algorithms are suitable to process large amount of data and tion success using only 3 features based on central statistics.
consistently derived the thresholds value or even fully develop Nevertheless, these results needed to be evaluated based on
improved algorithms. Furthermore, these have adaptive capa- the characteristics of the training and testing data which
bility that can be applied to particularize to specific drivers. proceeded from the same route, season and analogous traffic
These beneficial characteristics have promoted research into conditions [43].
data-driven approaches and is conditioning future trend in An Elman-type neural network (NN) was applied by
DS recognition. Proof of that is found in the extensive Augustynowicz to identify DS using speed and throttle pedal
literature available and represented in the following with a position [51]. Another application of NN was proposed by
predominant coverage with respect to RB and model-based Xu et al. [45] who developed a distal learning control frame-
approaches. work with two feedforward networks. NN1 concerned the
1) Unsupervised Machine Learning Algorithms: Unsuper- plant inverse model and NN2 simulated the plant forward
vised algorithms do not require understanding of the underly- model, which was run in parallel for error detection and self-
ing process. The classification is achieved through statistical learning. A comparison between various supervised learning
analysis of the input signals inherent to the algorithms. This algorithms was performed by Karginova et al. [30] who
allows increasing the number of classes and therefore the implemented: kNN, NN, decision tree and random forest.
finesse of the classification. Constantinescu et al. [17] pro- DS recognition using kNN achieved its best performance
posed two alternative algorithms based on hierarchical cluster within nearest neighbour inherent limits when clustering for
analysis and principal components analysis. The first one iden- k = 4 or 5. NN optimal number of hidden nodes was obtained
tifies groups of drivers based on similarities in the driving sta- as 20 with two approaches. The first network was trained
tistics, whilst the second option studies the correlation between with unmodified data, whilst the second was trained twice
variables to transform the original data set into a smaller with normal training data to encourage the preference for
one, whilst keeping the useful information. Each algorithm this style. As expected, the second NN showed less incorrect
returned a different number of classes, 5 and 7 in principal classification of normal DS. Random forest implementation
components and hierarchical clustering respectively [17]. The was performed to improve the poor performance of a single
use of principal components with extensive analysis of the decision tree. The tests revealed an optimal number of 40 trees,
driving-related extracted features is included in [37]. which achieved the best results in simulation environment.
Miyajima et al. [23] implemented a Gaussian Mixture Again the algorithm performance in normal style classification
Model (GMM) based on car-following behaviour and pedals was high, but returned unsatisfactory results in aggressive style
operation spectral analysis. The model achieved an identifi- classification. The algorithms assessment showed applicability
cation rate of 69% car-following, whilst the pedal spectral in real-time classification in all cases, but preference for
analysis provided a classification rate of 89.6% in simulation NN candidates based on the classification performance [30].
environment and 76.8% in field tests [23]. A similar applica- Markov models have also been successfully implemented
tion of GMM for pedal operation modelling was also reviewed for DS recognition. In Markov chains, future states only
by Wang and Lukic in [25]. Johnson and Trivedi applied depend on the current states, whose transition is generally
dynamic time wrapping to detect aggressive events with a 97% described by the so-called transition probability matrix [54].
success in simulation environment [14]. The stochastic and random characteristics of driver behav-
Driving style statistical characteristics also encourages the iour can be captured using Markov models as claimed by
use of Bayesian methods. Mudgal et al. [27] implemented a Guardiola et al. [33], who defended the benefits of Markov
hierarchical Bayesian regression model to characterize DS in Models combined with Monte Carlo application to generate
roundabouts. Similarly, McCall and Trivedi [40] employed random patterns based on past data in representation of DS.
Bayesian learning to assess critical situations related to braking Pentland and Liu defended Dynamical Markov Model suitabil-
assistance. The model was used to assess when a braking ity to capture DS, supported by the fact that human actions are
would be required and the probability of the driver performing best captured as a sequence of control steps [55]. Similarly,
this manoeuvre [40]. a Hidden Markov Model was used by Nechyba and Xu to
Unsupervised algorithms have proved their applicability and learn human control and validate the results [56].
suitability for DS recognition. Nevertheless, the output needs 3) Combined Unsupervised and Supervised Algorithms:
to be directed in terms of interpretation and number of clusters. Although supervised and unsupervised learning could be inter-
Besides, the classification performance can be worse compared preted as alternative methods, these can be combined in two-
to supervised algorithms. part algorithms to benefit from their advantages and improve
2) Supervised Machine Learning Algorithms: Supervised the overall performance. Unsupervised learning contributes to
algorithms imply knowledge of the driving style of the data identify the relevant input signals and the most efficient clas-
used for training. One of the simplest supervised algorithms sification strategy, whilst supervised learning generally returns
Authorized licensed use limited to: Universitaetsbibliothek der RWTH Aachen. Downloaded on May 02,2021 at 19:16:50 UTC from IEEE Xplore. Restrictions apply.
MARINA MARTINEZ et al.: DRIVING STYLE RECOGNITION FOR INTELLIGENT VEHICLE CONTROL AND ADVANCED DRIVER ASSISTANCE 673
better classification results. Bolovinou et al. [35] provided a Syed et al. in [15] and [58], who analysed drivers’ response
survey of data-driven algorithms combining supervised and acceptance to the feedback and adjusted it to reduce its intru-
unsupervised methods based on Bayesian inference, GMM, siveness to less receptive users. Reichart et al. [59] described a
Support Vector Machine (SVM), K-mean clustering, Bayesian similar approach, by increasing gradually the resistance over
Network and Markov models. However no concluding compar- the throttle pedal when the vehicle operation diverges from
ison was delivered due to the incompatibility caused by differ- optimal use adaptive to the driver response to the feedback.
ences in the input data and DS definition [35]. Van et al. [42] Alternative approaches are especially designed to improve
explored the application of K-means along with SVM. safety and efficiency in larger vehicles such as busses and
K-means unsupervised algorithm was used to explore the fleet management [17], [26], [43] and even passengers comfort
relevance of specific manoeuvres and their combination along in public transport [17]. Drivers of these means of trans-
with the driver pedals operation, whilst SVM was used for port could benefit from a personalized driving assistance,
implementation. “Driving coach”, as described by Gilman et al. [39], system
provided by context awareness to adapt to specific routes.
V. I NTELLIGENT V EHICLES AND C ONTROL A PPLICATIONS Rossetti et al. [60] proposed the use of serious game in trans-
DS recognition has several applications such as: driving portation as educational technique to improve driving skills in
style feedback and correction, driver-in-the-loop, ADASs per- simulators by analysing behaviour assimilation, elicitation and
formance enhancement and energy management in HEV, persuasion. Future trends of driver feedback will probably lean
PHEV and BEV. These can be gathered in two main groups, to a personalized assistance, where the system will be able to
safety or fuel efficiency aware. These alternative classifications detect the driver preferences and generate consistent feedback
can correlate at times given the fact that aggressive driving to both encourage style correction and avoid driver annoyance.
is generally associated to inefficiency, but do not coincide Ideally, this guidance should combine haptic and passive
in all cases. That is, calm drivers do not necessarily drive feedback with appropriate amount of information easily to
efficiently. interpret and follow, based on the driver cognitive capacity
and skills.
A. Driver Advisory Feedback
One of the main applications of DS recognition targets B. Enhance of ADAS Performance
driver style correction through either online or offline feed-
As previously mentioned, safety related systems would
back, which can be passive, active or a combination of
benefit from knowledge about DS to predict and anticipate
both. Corrective systems can be directly installed in the
to drivers’ reactions and adjust to individual users. ADASs
vehicle or included in smartphones to make use of the device
adaptation to drivers may result in safety systems of better
sensors and processing capability [13], [14], [18]. An example
performance [25], [34]. This would be the case of colli-
of DS recognition and correction using smartphones was pre-
sion avoidance systems that include driver-in-the-loop, which
sented by Stoichkov. The application was designed to display
would be able to anticipate to drivers’ reactions and actuate
visual feedback to promote DS improvement for both fuel con-
accordingly [9], [40]. Nonetheless, implementing ADASs in
sumption and risks of accidents reduction [18]. An alternative
driver-adaptive manner is a complex task that requires further
approach was proposed by Fazeen et al. who generated instead
investigation in the appropriate way of using DS information
audio feedback to encourage DS correction [11], as is similarly
in intelligent vehicles [35]. Kurz et al. [50] presented a method
used in [14]. Corti et al. [13] designed an algorithm with
to classify and use DS to improve the vehicle handling based
three alternative levels of visual information tested against the
on an online/offline algorithm updated with historical data.
no-feedback scenario. All approaches showed improvement
The authors also propose person-specific implementation for
in the vehicle consumption up to 20-30%, although best
multiple drivers [50].
results were achieved when providing enough information in
Additional safety-related applications can be found in coop-
a clearest and most understandable manner [13]. Doshi and
erative driving frameworks such as tandem and platoon,
Trivedi went a step further and analysed the response of drivers
as exposed by Rass et al. [61]. The system assumed the drivers
to the feedback to adapt the information facilitated and obtain
were willing to cooperate and provided feedback to encourage
individual evaluation [34].
safety and improve the overall fleet efficiency [61]. Similar
Although passive feedback strategies are easy to imple-
ideas were reviewed by Fotouhi et al. [62] to reduce fuel
ment and can return satisfactory results, it always exists the
consumption and emissions. Safe autonomous driving can be
possibility that the driver chooses to ignore the guidance.
an ultimate application of DS throughout increasing levels of
In contrast, active feedback systems generate a response that
automation. Urmson et al. [63] proposed an algorithm able
affects the normal vehicle operation, which cannot be as easily
to learn from drivers during manual operation and reproduce
neglected. Syed et al. provided both visual and haptic feedback
human-like manoeuvres autonomously within pre-established
registering up to 3.5% improvement in fuel economy without
safety constraints.
reducing the expectations of vehicle performance as claimed
by the authors. Haptic feedback was generated through the
throttle pedal [45], [57]. Nevertheless, haptic guidance can C. Control of HEV, PHEV and BEV
easily cause the driver annoyance, which could lead to the Although conventional vehicles are strongly affected by DS,
system manual deactivation. This issue was palliated by hybrid and battery electric vehicles are less sensitive to it
Authorized licensed use limited to: Universitaetsbibliothek der RWTH Aachen. Downloaded on May 02,2021 at 19:16:50 UTC from IEEE Xplore. Restrictions apply.
674 IEEE TRANSACTIONS ON INTELLIGENT TRANSPORTATION SYSTEMS, VOL. 19, NO. 3, MARCH 2018
due to the higher efficiency of the electric components, idling VI. C ONCLUDING R EMARKS
reduction and regenerative braking. Lenaers [46] asserted the This article offers a review of the recent research and devel-
influence of DS and road type combinations on consumption opment efforts on driving style characterization and recogni-
and CO2 emissions in four different powertrains: petrol, diesel, tion as well as their applications to intelligent vehicle control.
hybrid and liquefied petroleum gas. The hybrid powertrain The design process is chronologically followed from input
was able to provide lower consumption even under urban signals identification and classification policy definition to
aggressive drive, particularly due to start-stop and regeneration the algorithm selection and implementation. All driving style
during braking. Furthermore, aggressive driving styles was influencing factors and classification strategies are presented
found to require up to 68% and 47% more fuel in urban in relationship to the targeted applications and implementation
and rural roads respectively in conventional powertrains, whilst constraints. The complexity of driving style is discussed and
liquid petroleum and most importantly hybrid powertrains are the predominant interpretations, safety and fuel efficiency-
less sensitive to the different driving patterns [46]. Neubauer related, are analysed through the enumeration of different
and Wood compared the effect of DS on a conventional algorithms. The gradual increment of ADAS presence and
vehicle, HEV, PHEV and BEV in terms of optimal operating vehicle autonomous capabilities requires deeper driving style
conditions. The results showed important influence of driving analysis and the inclusion of drivers in the systems. This has
style in all vehicle platforms, although this was affecting dif- promoted the use of data-driven algorithms able to process
ferently. HEV presented optimal working points at low speed, more data and the implementation of machine learning algo-
whilst conventional vehicles improved the overall efficiency at rithms able to adapt to individual drivers. Future trends will
higher speed values. Furthermore, HEVs were barely affected focus on even larger data sets covering a broader plurality
by braking due to the regenerative capability, but particularly of drivers and will implement the combination of supervised
vulnerable to high acceleration [48]. A part from different and unsupervised algorithms to enhance driver adaptability
operating conditions, hybrid modes selection in HEV and and cognitive performance. The presented review positions
PHEV are strongly related with the driver torque request and driving style recognition as a crucial concept for intelligent
inevitably to DS. Consequently, slight changes in driving style vehicle development that will strongly condition the progress
can cause unnecessary modes switching and develop sub- in autonomous vehicles and their integration in the market.
optimal performance [15].
Besides, energy management can be enhanced with prior R EFERENCES
knowledge of driving style in HEV and PHEV. The increas- [1] F.-Y. Wang, “Parallel control and management for intelligent transporta-
ing level of vehicle electrification has motivated research tion systems: Concepts, architectures, and applications,” IEEE Trans.
Intell. Transp. Syst., vol. 11, no. 3, pp. 630–638, Sep. 2010.
on the effect of DS and driving events on the energy [2] L. Li, D. Wen, N.-N. Zheng, and L.-C. Shen, “Cognitive cars: A new
demand in PHEVs, being both contributions particularly frontier for ADAS research,” IEEE Trans. Intell. Transp. Syst., vol. 13,
no. 1, pp. 395–407, Mar. 2012.
important for optimal control [12]. An optimal control strat- [3] J. Zhang, F.-Y. Wang, K. Wang, W.-H. Lin, X. Xu, and C. Chen, “Data-
egy can be obtained with knowledge of the entire driving driven intelligent transportation systems: A survey,” IEEE Trans. Intell.
cycle, environment and driver behaviour. Thus the infor- Transp. Syst., vol. 12, no. 4, pp. 1624–1639, Dec. 2011.
[4] Y. Dong, Z. Hu, K. Uchimura, and N. Murayama, “Driver inattention
mation about driving style can potentially lead to impor- monitoring system for intelligent vehicles: A review,” IEEE Trans. Intell.
tant fuel savings [25]. Driving style information was used Transp. Syst., vol. 12, no. 2, pp. 596–614, Jun. 2011.
by Lin et al. [64] to improve the energy management of [5] T. Qu, H. Chen, D. Cao, H. Guo, and B. Gao, “Switching-based
stochastic model predictive control approach for modeling driver steering
a hybrid truck. Yu et al. [65] used driving pattern and skill,” IEEE Trans. Intell. Transp. Syst., vol. 16, no. 1, pp. 365–375,
style recognition to better estimate the remaining range Feb. 2015.
[6] Y. Zheng, S. Li, J. Wang, D. Cao, and K. Li, “stability and scalability of
in PHEV assuming static and quasi-static knowledge of homogeneous vehicular platoon: Study on the influence of information
the route. Similarly, Zhang et al. [66] proposed an adap- flow topologies,” IEEE Trans. Intell. Transp. Syst., vol. 17, no. 1,
tive control strategy for a PHEV using driving pattern pp. 14–26, Jan. 2016.
[7] K. Bengler, K. Dietmayer, B. Farber, M. Maurer, C. Stiller, and
recognition. H. Winner, “Three decades of driver assistance systems: Review and
Driving style recognition can also provide useful infor- future perspectives,” IEEE Intell. Transp. Syst. Mag., vol. 6, no. 4,
mation to estimate more accurately the remaining range, pp. 6–22, Oct. 2014.
[8] F.-Y. Wang, “Agent-based control strategies for smart and safe vehicles,”
improve range management and potentially extend it [22]. in Proc. IEEE Int. Conf. Veh. Electron. Safety, Oct. 2005, pp. 331–332.
The effect of DS over the BEV range was highlighted by [9] F.-Y. Wang, “Scanning the issue and beyond: Parallel driving with
software vehicular robots for safety and smartness,” IEEE Trans. Intell.
Bingham et al. [67], who claimed a possible divergence of Transp. Syst., vol. 15, no. 4, pp. 1381–1387, Aug. 2014.
10% between drivers. This margin was associated to aggres- [10] W. Barfield and T. A. Dingus, Human Factors in Intelligent Trans-
sive deceleration events that triggered friction brakes [67]. portation Systems. Virginia Polytech. Inst. State Univ., Psychol. Press,
Blacksburg, VA, USA, 2014.
Similarly, Bolovinou et al. [24] developed a strategy able to [11] T. Lee and J. Son, “Relationships between driving style and fuel
estimate the remaining range using route information: speed, consumption in highway driving,” SAE Tech. Paper 2011-28-0051, 2011.
[12] B. V. Padmarajan, A. McGordon, and P. A. Jennings, “An investigation
gradient profile, traffic conditions and DS. This study targeted on the effect of driver style and driving events on energy demand of a
an accurate estimation of the battery range to reduce the PHEV,” in Proc. Electric Veh. Symp., Los Angeles, CA, USA, May 2012,
so-called range anxiety that characterises BEV drivers and is pp. 1–9.
[13] A. Corti, C. Ongini, M. Tanelli, and S. M. Savaresi, “Quantitative driving
found to be the main constraint that hinders electric vehicles style estimation for energy-oriented applications in road vehicles,” in
integration in the market [24]. Proc. IEEE Int. Conf. Syst., Man, Cybern., Oct. 2013, pp. 3710–3715.
Authorized licensed use limited to: Universitaetsbibliothek der RWTH Aachen. Downloaded on May 02,2021 at 19:16:50 UTC from IEEE Xplore. Restrictions apply.
MARINA MARTINEZ et al.: DRIVING STYLE RECOGNITION FOR INTELLIGENT VEHICLE CONTROL AND ADVANCED DRIVER ASSISTANCE 675
[14] D. A. Johnson and M. M. Trivedi, “Driving style recognition using a [38] D. Dörr, D. Grabengiesser, and F. Gauterin, “Online driving style
smartphone as a sensor platform,” in Proc. IEEE Conf. Intell. Transp. recognition using fuzzy logic,” in Proc. IEEE 17th Int. Conf. Intell.
Syst., Oct. 2011, pp. 1609–1615. Transp. Syst., Oct. 2014, pp. 1021–1026.
[15] F. Syed, S. Nallpa, A. Dobryden, C. Grand, R. McGee, and D. Filev, [39] E. Gilman, A. Keskinarkaus, S. Tamminen, S. Pirttikangas, J.
“Design and analysis of an adaptive real-time advisory system for Röning, and J. Riekki, “Personalised assistance for fuel-efficient
improving real world fuel economy in a hybrid electric vehicle,” SAE driving,” Transp. Res. C, Emerg. Technol., vol. 58, pp. 681–705,
Tech. Paper 2010-01-0835, 2010. Sep. 2015.
[16] M. Fazeen, B. Gozick, R. Dantu, M. Bhukhiya, and M. C. Gonzalez, [40] J. C. McCall and M. M. Trivedi, “Driver behavior and situation aware
“Safe driving using mobile phones,” IEEE Trans. Intell. Transp. Syst., brake assistance for intelligent vehicles,” Proc. IEEE, vol. 95, no. 2,
vol. 13, no. 3, pp. 1462–1468, Sep. 2012. pp. 374–387, Feb. 2007.
[17] Z. Constantinescu, C. Marinoiu, and M. Vladoiu, “Driving style analysis [41] H. Xia, K. Boriboonsomsin, and M. Barth, “Dynamic eco-driving
using data mining techniques,” Int. J. Comput., Commun. Control, vol. 5, for signalized arterial corridors and its indirect network-wide
no. 5, pp. 654–663, Dec. 2010. energy/emissions benefits,” J. Intell. Transp. Syst., vol. 17, no. 1,
[18] R. Stoichkov, “Android smartphone application for driving style recog- pp. 31–41, Jan. 2013.
nition,” Ph.D. dissertation, Dept. Elect. Eng. Inf. Technol, Technische [42] M. Van Ly, S. Martin, and M. M. Trivedi, “Driver classification and
Univ. München, Munich, Germany, 2013. driving style recognition using inertial sensors,” in Proc. IEEE Intell.
[19] E. Ericsson, “Variability in urban driving patterns,” Transp. Res. Part D, Veh. Symp. (IV), Jun. 2013, pp. 1040–1045.
Transp. Environ., vol. 5, no. 5, pp. 337–354, 2000. [43] V. Vaitkus, P. Lengvenis, and G. Z̆ylius, “Driving style classification
[20] E. Ericsson, “Independent driving pattern factors and their influence on using long-term accelerometer information,” in Proc. 19th IEEE Int.
fuel-use and exhaust emission factors,” Transp. Res. D, Transp. Environ., Conf. Methods Models Autom. Robot., Sep. 2014, pp. 641–644.
vol. 6, no. 5, pp. 325–345, 2001. [44] H. Xiong and L. N. Boyle, “Drivers’ adaptation to adaptive cruise
[21] Y. L. Murphey, R. Milton, and L. Kiliaris, “Driver’s style classification control: Examination of automatic and manual braking,” IEEE Trans.
using jerk analysis,” in Proc. IEEE Workshop Comput. Intell. Vehicles Intell. Transp. Syst., vol. 13, no. 3, pp. 1468–1473, Sep. 2012.
Veh. Syst., Mar./Apr. 2009, pp. 23–28. [45] L. Xu, J. Hu, H. Jiang, and W. Meng, “Establishing style-oriented
[22] J. Neubauer and E. Wood, “Thru-life impacts of driver aggression, driver models by imitating human driving behaviors,” IEEE Trans. Intell.
climate, cabin thermal management, and battery thermal manage- Transp. Syst., vol. 16, no. 5, pp. 2522–2530, Oct. 2015.
ment on battery electric vehicle utility,” J. Power Sour., vol. 259, [46] G. Lenaers, “Real life CO2 emission and consumption of four
pp. 262–275, Aug. 2014. car powertrain technologies related to driving behaviour and
[23] C. Miyajima et al., “Driver modeling based on driving behavior and road type,” in Proc. 9th Int. Conf. Engines Veh., Sep. 2009,
its evaluation in driver identification,” Proc. IEEE, vol. 95, no. 2, pp. 1–15.
pp. 427–437, Feb. 2007. [47] F. U. Syed, D. Filev, and H. Ying, “Fuzzy rule-based driver advisory
[24] A. Bolovinou, I. Bakas, A. Amditis, F. Mastrandrea, and W. Vinciotti, system for fuel economy improvement in a hybrid electric vehicle,” in
“Online prediction of an electric vehicle remaining range based on Proc. Annu. Meeting North Amer. Fuzzy Inf. Process. Soc., Jun. 2007,
regression analysis,” in Proc. IEEE Int. Elect. Veh. Conf., Dec. 2014, pp. 178–183.
pp. 1–8. [48] J. Neubauer and E. Wood, “Accounting for the variation of driver
[25] R. Wang and S. M. Lukic, “Review of driving conditions prediction and aggression in the simulation of conventional and advanced vehicles,”
driving style recognition based control algorithms for hybrid electric SAE Tech. Paper 2013-01-1453, 2013.
vehicles,” in Proc. IEEE Veh. Power Propuls. Conf., Sep. 2011, pp. 1–7. [49] J. Huang, Y.-K. Chin, and W. C. Lin, “Adaptive vehicle control system
[26] V. Manzoni, A. Corti, P. De Luca, and S. M. Savaresi, “Driving with driving style recognition,” U.S. Patent 8 260 515, Sep. 4, 2012.
style estimation via inertial measurements,” in Proc. IEEE Conf. Intell. [50] G. Kurz, A. Müller, T. Röhring-Gericke, R. Schöb, H. Tröster, and
Transp. Syst., Sep. 2010, pp. 777–782. A. Yap, “Method and device for classifying the driving style of a driver
[27] A. Mudgal, S. Hallmark, A. Carriquiry, and K. Gkritza, “Driving in a motor vehicle,” U.S. Patent 6 449 572, Sep. 10, 2012.
behavior at a roundabout: A hierarchical Bayesian regression analysis,” [51] A. Augustynowicz, “Preliminary classification of driving style with
Transp. Res. D, Transp. Environ., vol. 26, pp. 20–26, Jan. 2014. objective rank method,” Int. J. Autom. Technol., vol. 10, no. 5,
[28] O. Taubman-Ben-Ari, M. Mikulincer, and O. Gillath, “The multidimen- pp. 607–610, 2009.
sional driving style inventory—Scale construct and validation,” Accident [52] J.-S. Won, “Intelligent energy management agent for a parallel hybrid
Anal. Prevention, vol. 36, no. 3, pp. 323–332, 2004. vehicle,” Ph.D. dissertation, Texas A&M University in partial fulfilment
[29] (2010). Exponent Engineering & Scientific Consulting, accessed on of the requirements for the degree of doctorate of philosophy, Pusan
May 16, 2015. [Online]. Available: http://www.exponent.com/ Nat. Univ., Busan, South Korea, May 2003.
human_factors_vehicle/ [53] J. Kim, H. Sim, and J. Oh, “The flexible EV/HEV and SOC band control
[30] N. Karginova, S. Byttner, and M. Svensson, “Data-driven methods for corresponding to driving mode, driver’s driving style and environmental
classification of driving styles in buses,” SAE Tech. Paper 2012-01-0744, circumstances,” SAE Tech. Paper 2012-01-1016, 2012.
2012. [54] W.-K. Ching, X. Huang, M. K. Ng, and T. K. Siu, Markov Chains.
[31] D. Filev, J. Lu, K. Prakah-Asante, and F. Tseng, “Real-time driving Models, Algorithms and Applications, International Series in Operations
behavior identification based on driver-in-the-loop vehicle dynamics Research & Management Science, vol. 189, 2nd ed. New York, NY,
and control,” in Proc. IEEE Int. Conf. Syst., Man, Cybern., Oct. 2009, USA: Springer, 2013.
pp. 202–220. [55] A. Pentland and A. Liu, “Modeling and prediction of human behaviour,”
[32] W. Wang, J. Xi, and H. Chen, Modeling and Recognizing Driver Neural Comput., vol. 11, no. 1, pp. 229–242, 1999.
Behavior Based on Driving Data: A Survey. Cairo, Egypt: Hindawi [56] M. C. Nechyba and Y. Xu, “On the fidelity of human skill models,” in
Publishing Corporation, 2014. Proc. IEEE Int. Conf. Robot. Autom., vol. 3. Apr. 1996, pp. 2688–2693.
[33] C. Guardiola, B. Pla, D. Blance-Rodriguez, and A. Reig, “Modelling [57] F. U. Syed and D. Filev, “Real time advisory system for fuel economy
driving behaviour and its impact on the energy management problem improvement in a hybrid electric vehicle,” in Proc. Annu. Meeting North
in hybrid electric vehicles,” Int. J. Comput. Math., vol. 91, no. 1, Amer. Fuzzy Inf. Process. Soc., May 2008, pp. 1–6.
pp. 147–156, 2014. [58] F. U. Syed, D. Filev, F. Tseng, and H. Ying, “Adaptive real-time advisory
[34] A. Doshi and M. M. Trivedi, “Examining the impact of driving style system for fuel economy improvement in a hybrid electric vehicle,” in
on the predictability and responsiveness of the driver: Real-world Proc. Annu. Meeting North Amer. Fuzzy Inf. Process. Soc., Jun. 2009,
and simulator analysis,” in Proc. IEEE Intell. Veh. Symp., Jun. 2010, pp. 1–7.
pp. 232–237. [59] G. Reichart, S. Friedmann, C. Dorrer, H. Rieker, E. Drechsel, and
[35] A. Bolovinou, F. Bellotti, A. Amditis, and M. Tarkiani, “Driving G. Wermuth, “Potentials of BMW driver assistance to improve fuel econ-
style recognition for co-operative driving: A survey, ADAPTIVE omy,” in Proc. FISITA World Autom. Congr., Paris, France, Sep. 1998,
2014,” in Proc. 6th Int. Conf. Adapt. Self-Adapt. Syst. Appl., 2014, vol. 27, pp. 1–16.
pp. 73–78. [60] R. J. F. Rossetti, J. E. Almeida, Z. Kokkinogenis, and J. Goncalves,
[36] X. Huang, Y. Tan, and X. He, “An intelligent multifeature statistical “Playing transportation seriously: Applications of serious games to
approach for the discrimination of driving conditions of a hybrid electric artificial transportation systems,” IEEE Intell. Syst., vol. 24, no. 4,
vehicle,” IEEE Trans. Intell. Transp. Syst., vol. 12, no. 2, pp. 453–465, pp. 104–112, Jul./Aug. 2013.
Jun. 2011. [61] S. Rass, S. Fuchs, and K. Kyamakya, “A game-theoretic approach to co-
[37] M. Ishibashi, M. Okuwa, S. Doi, and M. Akamatsu„ “Indices for char- operative context-aware driving with partially random behavior, Smart
acterizing driving style and their relevance to car following behavior,” Sens. Context, Lecture Notes Comput. Sci., vol. 5279, pp. 154–167,
in Proc. Annu. Conf. SICE, Sep. 2007, pp. 1132–1137. 2008.
Authorized licensed use limited to: Universitaetsbibliothek der RWTH Aachen. Downloaded on May 02,2021 at 19:16:50 UTC from IEEE Xplore. Restrictions apply.
676 IEEE TRANSACTIONS ON INTELLIGENT TRANSPORTATION SYSTEMS, VOL. 19, NO. 3, MARCH 2018
[62] A. Fotouhi, R. Yusof, R. Rahmani, S. Mekhilef, and N. Shateri, Bo Gao received the Ph.D. degree in mechani-
“A review on the applications of driving data and traffic information for cal engineering from the University of Bath, U.K.
vehicles’ energy conservation,” Renew. Sustain. Energy Rev., vol. 37, He was a Research Officer with the Department of
pp. 822–833, Sep. 2014. Mechanical Engineering, University of Bath, where
[63] C. Urmson, D. Dolgov, and P. Nemec, “Driving pattern recognition and he was involved in vehicle dynamics and stability
safety control,” U.S. Patent 8 634 980, Jan. 21, 2014. control. His Ph.D. research was on the development
[64] C.-C. Lin, S. Jeon, H. Peng, and J. M. Lee, “Driving pattern recognition of a novel hydro-pneumatic active suspension system
for control of hybrid electric trucks,” Int. J. Veh. Mech. Mobility, vol. 42, supported by Jaguar Land Rover and Ford Motor
nos. 1–2, pp. 41–58, 2004. Company. Since joining AVL in 2006, he has been
[65] H. Yu, F. Tseng, and R. McGee, “Driving pattern identification for involving in various vehicle programs involved with
EV range estimation,” in Proc. IEEE Int. Elect. Veh. Conf., Mar. 2012, several OEMs and Tier 1 suppliers. He is currently
pp. 1–7. the Department Manager of the Electronics & Controls Division, AVL
[66] S. Zhang and R. Xiong, “Adaptive energy management of a plug-in Powertrain UK Ltd. He is currently leading the vehicle electrification and
hybrid electric vehicle based on driving pattern recognition and dynamic
ADAS activities within the company. He holds one patent and has co-authored
programming,” Appl. Energy, vol. 155, pp. 68–78, Oct. 2015. over 20 technical papers. He is also a member of the AVL Global Controls
[67] C. Bingham, C. Walsh, and S. Carroll, “Impact of driving characteristics
on electric vehicle energy consumption and range,” IET Intell. Transp. Committee. He is also a Chartered Engineer and a member of the Institution
Syst., vol. 6, no. 1, pp. 29–35, 2012. of Mechanical Engineers.
Authorized licensed use limited to: Universitaetsbibliothek der RWTH Aachen. Downloaded on May 02,2021 at 19:16:50 UTC from IEEE Xplore. Restrictions apply.