See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/305852225

Vehicle gear shifting strategies co-simulations to optimize performance and

fuel consumption at high speeds and accelerations

Conference Paper · January 2014

CITATIONS READS

0 271

5 authors, including:

Jony Eckert Fabio Santiciolli

University of Campinas University of Campinas
69 PUBLICATIONS   160 CITATIONS    49 PUBLICATIONS   129 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Development of a Robust Control System for Emulating Real Driving Conditions in a Roller Chassis Dynamometer View project

Multibeam vPEH Optimization View project

All content following this page was uploaded by Jony Eckert on 05 August 2016.

The user has requested enhancement of the downloaded file.

 Vehicle gear shifting strategies co-simulations to optimize performance and fuel
consumption at high speeds and accelerations.

Jony Javorski Eckert

Fabio Mazzariol Santiciolli
Eduardo dos Santos Costa
Heron José Dionísio
Franco Giuseppe Dedini
State university of Campinas-UNICAMP

ABSTRACT

The driver gear shifting strategy influences significantly in the vehicle dynamic behavior, performance and fuel
consumption because it changes the transmission system inertia and the engine speed. In this paper was created an
algorithm that can merge strategies that maximize vehicle acceleration performance with others that aim fuel economy
according to the power demand. The analyses were performed through co-simulation between the multibody dynamics
TM TM
program Adams and Simulink/Matlab . The power demand was defined based on the motion resistance equations of
the vehicle longitudinal dynamics and the standard driving cycle US06 which provides the speeds profile comparison of
for simulations results.

INTRODUCTION

The vehicle dynamics studies the interactions between the vehicle, the driver and the environment. The literature
proposes to divide the vehicular dynamic into three areas: longitudinal, lateral and vertical. The longitudinal dynamics
evaluates the vehicle power required for remain the desired speed, evaluating by means of equations the forces acting on
the system. The power demand is a function of the aerodynamic drag, tire-ground interaction, climbing resistance,
powertrain inertia and driving behavior.

According to Wong (2001) the vehicle power is provided by the engine that can only be used with appropriate
transmission because of its performance characteristics. The powertrain is usually composed by clutch, gearbox, drive
shafts and differential. The gearbox provides a number of gear ratios selected to provide the: vehicle traction force to keep
the speed closer to the ideal.

The shift control governs the transmission behavior, during a shift event, in function of the resistance forces and vehicle
speed. The top gear is selected to get a maximum speed and it is limited by the engine power, speed and the fuel
economy (Singh et al., 2012). The longitudinal vehicle dynamics models proposed in the literature do not define a
standard gear shifting, due because this factor is represented by the driver behavior, depending on road and traffic
conditions. According to Kahlbau et al., (2013) efficiency and shift quality are determined by the shift strategy, defining at
what time a shift event is executed, and presents significant effects over the vehicular performance and fuel consumption
(Santiciolli et al., 2013).

Vagg et al. (2012), describes that the use of gear shifting indicators, composed of light and/or sound alarms connected to
management software installed in the Electronic Control Unit (ECU), results in 3.6% of fuel economy in a vehicle
submitted to the New European Driving Cycle. Furthermore, Gao et al. (2011) optimized the gear shift time in heavy
vehicles as a function of the engine torque and clutch performance. Guan and Frey (2012) developed a driving assistance
system to help drivers giving them fuel efficiency guidelines that are suited for the current situation of the vehicle. With the
same purpose of reduce fuel consumption Orfila et al. (2012) evaluated experimentally the variability of driver behaviors
related, in relation with the engine speeds amplitude according to the driving style for 21 drivers. After that he reproduced
the tests using ecodriving techniques to reduce the fuel consumption.

The objective of this paper is evaluate the performance and fuel consumption of a 1.0L vehicle, based on US06 standard
TM TM
driving cycle, by co-simulations between Adams where is located the multibody vehicle model and Matlab/Simulink
where is implemented the vehicle dynamics equations and the gear shifting control strategies used.
LONGITUDINAL VEHICLE DYNAMICS

In this paper it will be used the longitudinal vehicle dynamics methodology proposed by Gillespie (1992) where the model
is based on the acting forces on the vehicle travel direction as shown in Figure 1.

Figure 1 - Arbitrary forces acting on a vehicle (Gillespie, 1992)

AERODYNAMIC DRAG

The aerodynamic load ( ) is the resistance imposed by the air during the vehicle passage. This effect is proportional to
the square of the vehicle speed. According to Ehsani et al. (2009), a vehicle traveling at a particular speed in air,
generates a resistance force of its motion. This force is known as aerodynamic drag and it is resultant from two
components: shape drag and skin friction. Due to the complexity of the airflow outside the vehicle, this load is based on
empirical constant and a term known as drag coefficient, as shown in Eq. (1).

(1)

ROLLING RESISTANCE

Rolling resistance is a result of energy loss in the tire, which can be traced back to the deformation of the area of tire
contact and the damping properties of the rubber. These lead to the transformation of mechanical into thermal energy,
contributing to warming of the tire (Reimpell and Stoll, 1996). At low speeds on hard pavement, rolling resistance ( ) is
the primary resistance load caused essentially by: the tire deformation, the pavement and the tire adhesion on the ground.
This paper will consider hard surfaces, such as asphalt and concrete. In these cases, the ground stiffness is higher than
the tires, therefore the road can be considered undeformable. The rolling resistance is shown by the Eq. (2).

(2)

Where is the vehicle weight [ ] and represents the rolling resistance coefficient calculated by the Eq. (3) in function
of the vehicle speed .

( ) (3)

ROAD GRADE INFLUENCE

This term refers to the weight force decomposition resulting from the road grade. In uphill, the weight force component
acts retarding the vehicle movement, and in downhill, the weight force aids the movement. The grade angle also results in
a component of weight parallel to the ground which, as in the case of accelerating or braking on flat ground, results in a
longitudinal weight transfer. The effects of grade and longitudinal acceleration can be combined in finding the changes in
front and rear loads due to both (Milliken et al., 1995). The road grade is considered null in US06 cycle, therefore the
terms relating to the climbing resistance will be removed from the vehicle longitudinal dynamics equationing proposed.

ACCELERATION PERFORMANCE

The vehicle acceleration generates resistance forces as the vehicle longitudinal displacement as the powertrain rotational
inertia. The available traction force ( ) in function of the engine torque and the transmission ratio is given by Eq. (4).

(( ) ) (4)

• = Available engine torque [ ]; • = Gearbox inertia [ ];

• = Total gear ratio; • = Differential inertia [ ];

• = Transmission overall efficiency; • = Wheels and tires inertia [ ];

• = Tire external radius [ ]; • = Gearbox transmission ratio;

• = Gearbox inertia [ ]; • = Vehicle longitudinal acceleration [ ].

The vehicle acceleration performance is given by the Eq. (5), where is the vehicle mass [ ] and the road grade [rad].

( ) (5)

CO-SIMULATION

The longitudinal dynamics simulation is normally used to compare the importance of energy balance characteristics for
vehicles, without the need of building prototypes that can take much time and a high cost (Oliveira, 2005). The co-
simulation technique is used in a development stage where the physical or mathematical mechatronic and control system
are designed (Brezina et al., 2011).

The simulations implemented in this paper were made via the multibody dynamic analysis program Adams™ (Automatic
Dynamic Analysis of Mechanical Systems), where is implemented the vehicular model analyzed. The control of variables
related to longitudinal dynamics, as described earlier, is done through the interface between Adams™ and
MATLAB/Simulink™.

The simulated vehicle was based on a compact hatchback equipped with 1.0L engine (Table 1). The implemented model
was designed based on a dynamometer bench (Figure 2) to enable future experimental validations. The effects of vehicle
suspension system were neglected to simplify the model. These factors are also disregarded by the current literature.

Table 1 - Vehicle parameters

Speed
Components Units
1st 2nd 3rd 4th 5th
Engine inertia 0.1367
Transmission inertia 0.0017 0.0022 0.0029 0.0039 0.0054
Transmission ratio - 4.27 2.35 1.48 1.05 0.8
Diferencial inertia 9.22E-04
Diferencial ratio - 4.87
Wheels + tires inertia 2
Vehicle mass 980
Tires - 175/70 R13
 Figure 2 - CAD model

The CAD model was exported to Adams™ where an appropriate revolution joints were created to allow the wheels
movement and rotating masses. On the wheels were applied torques related to the powertrain and brake system in the
rotating masses were applied a movement resistance torque. In the model, the vehicle chassis was connected to the base
to prevent longitudinal movement so that the wheels remain aligned with the rollers. The rotational movement between the
rollers and the wheels are done by means of a joint, transmitting torques and acting speeds.

To facilitate the implementation of the vehicle dynamics equations, it was used a Simulink™/Adams™ interface,
generating a block of data from the dynamic model as shown in Figure 3.

Figure 3 - Adams™ generated block

The Simulink™ programmed algorithm works together with the Adams™ solver. The Simulink™ provides for Adams™
torque values applied in the front wheels. The Adams™ generates a response from an angular velocity of the wheels,
which supplies the Simulink™ algorithm to recalculate the required torque according to the new demand.
DRIVING CYCLE

With the intention to establish a benchmark, standard cycles were utilized to determine the vehicle speed behavior, in a
way that the mathematic model calculate the vehicle required power to follow the velocity profile predetermined by the
cycle. A driving cycle represents the way the vehicle is driven during a trip and the road characteristics. In the simplest
case, it is defined as a sequence of vehicle speed (and therefore acceleration) and road grade (Corrêa et al., 2011).
These driving cycles are designed to be representative of urban and extra-urban driving conditions, and reproduce
measures of vehicle speed in real roads. Some of them and the test procedures have been recently updated to better suit
modern vehicles, following criticism towards the previous regulation (Serrao et al., 2005).

In the simulations was used the US06 (Figure 4) cycle that complements the U.S. test cycle FTP-75 because it represents
high speeds and accelerations. This cycle represents a route of 12.8 , with an average speed of 77.9 , maximum
speed of 129.2 and duration of 596 seconds. Therefore, this cycle allows a better analysis of the vehicle
performance in acceleration condition, to evaluate a critical operation in the proposed model.

Figure 4 - Velocity profile US06

ENGINE PARAMETERS

The simulated model considers the engine torque curves Figure 5(A) that is the previously described mathematic model
compared with the available engine torque in function of the acceleration percentage and the engine speed. If the required
torque exceeds the maximum torque available, there will be loss of performance. The fuel consumption is given by the
map shown in Figure 5(B) as a function of torque and engine speed.

Figure 5 – (A) torque curve in function of the throttle. (B) Fuel consumption map
GEAR SHIFTING PARAMETERS

For a correct representation of the vehicle behavior is necessary to define some specific parameters to avoid gear shifting
instability. In the simulations performed was used a gear shifting time of 1 as proposed by Yin et al. (2007). Adwell time
between two subsequent gear shifts is also important for stabilization. This is required to avoid chattering and to satisfy
comfort conditions (Casavola et al., 2010). The downshift occurs at 5 km=h below the upshift speed, as proposed by Xi et
al. (2009), to prevent gear shift instability

GEAR SHIFTING STRATEGIES

Maximum power and torque strategies consist in optimize the vehicle acceleration performance, usually acting at engine
higher speeds, thereby generating a fuel consumption increase. The fuel economy strategy is more difficult to define
because it depends on a number of factors that vary from the engine behavior, available gear ratios, driver required
acceleration and fuel type. Taking into in account these parameters, the tactics consist in keeping the engine running in
higher efficiency regions. In this paper were used the gear shifting speeds proposed by GM (2013) for a similar vehicle
that the simulated. First it was observed that the fuel economy strategy upshift occurred when the engine speed
approached to 3000 , similarly the maximum torque strategy gear shifting close to the engine 5300 , and
maximum power tactic in 6400 . Due to the high engine speed range available between the fuel economy strategy
and the gear shifting at maximum engine torque, were included two intermediate strategies to make gear shifting at 3500
and 4500 engine .

RESULTS

Primarily simulations for the gear shifting tactics mentioned above were performed, the results were compared with the
standard US06 velocity profile, where it was observed that in high acceleration stretches, the vehicle did not have enough
power to keep the required speed, staying at lower speeds until the cycle achieve braking sections or lower acceleration
stretches.

LINEAR CORRELATION

To evaluate the difference between the simulation results, we used the linear correlation between the velocity profile and
the standard cycle adopted. The correlation coefficient is the intensity measure of the linear relationship between the
two variables. The term is the square of the correlation coefficient, called determination coefficient, and consists of the
sum of squares of prediction errors obtained as shown in Eq. (6), where and represent the curves values.

( ( ̅ )( ̅))
(6)
( ̅) ( ̅)

The regression measures the variability proportion between the two curves, therefore, is a direct correlation function
between the variables, showing the variance percentage of the variables. A value of close to 1 indicates a strong
relationship between the two variables.

SIMULATION RESULTS ANALYSIS

Due to the vehicle remain at speeds lower that the required by the standard velocity profile, the total distance traveled is
affected by the vehicle performance. This causes differences in the results, because in the simulations using fuel
economy strategies travel a shorter distance because of the performance decrease in high acceleration stretches, than
simulations using tactics focused on the vehicle acceleration performance, generating differences in the fuel consumption
average by traveled.
The results obtained by simulations of the gear shifting tactics are shown in Table 2. As the simulation objective is follow
the standard velocity profile imposed, correlations are always close to 1, and the differences were found in the third
decimal place.

Table 2 - Correlations and fuel consumption according to the gear shifting tactic

Fue lConsumption Traveled Distance Consumption Average

Linear
Gear Shifting Tactics ( ) ( ) ( )
Correlation
Fuel economy 0.99298 798.5 12.82 16.05

Gear Shifting at 3500 rpm 0.99467 806.1 12.84 15.93

Gear Shifting at 4500 rpm 0.99608 828.7 12.88 15.54

Maximum torque 0.99665 883.3 12.89 14.60

Maximum power 0.99703 971.9 12.90 13.27

Figure 6 shows the performance difference between the fuel economy and the maximum power strategies. Using the fuel
economy tactics the simulated vehicle does not follow the standard speeds on high acceleration stretches, because it has
not enough power to fulfill the cycle, however this tactic present the best Consumption Average, even the vehicle traveled
a lower distance in comparison with the standard velocity profile, because this strategy gives priority to engine better
efficiency regions.

Figure 6 - Comparison between simulated tactics and the speed US06 profile.
For the vehicle follow the standard velocity profile was necessary to use the engine maximum power gear shifting
strategy. This operation region is located at higher engine speeds, so the vehicle takes more time to reach the gear
shifting speed, staying for a longer period using gears with greater transmission ratio, resulting in a higher torque in the
wheels which increases the vehicle acceleration capacity. However this operation regime is located in an engine low
efficiency region generating a significant fuel consumption increase.

HYBRID STRATEGY

The driver does not keep a default strategy throughout all the route by simply adjusting the power required as needed (Yu
et al. 2004), for example, in the fuel economy strategy simulation, the vehicle could keep the required performance in a
large part of the cycle, only getting suboptimal performance in acceleration peaks.

The gear shifting tactics in manual transmissions should be assisted by an intelligent interactive system, as proposed by
Kang et al. (2012) using the gear shifting control to optimize the torque applied to the vehicle wheels. In order to maintain
the performance and optimize fuel consumption compared to the maximum torque and power strategies was developed
an adaptive algorithm that fits the gear shifting strategy as the power demand. This algorithm repeats the same simulation
few times analyzing the stretches in which the vehicle fails to achieve the desired speed by changing the gear shifting
strategy to achieve acceptable performance.

The hybrid strategy achieved a similar performance to maximum power gear shifting tactic, however with a fuel
consumption of 826.1 , representing 15% fuel economy compared to the maximum power strategy, and an increase of
3.45% over the fuel economy strategy, but with superior performance. The travelled distance was similar to the maximum
power strategy, which achieved the best performance between the simulated tactics, totaling a fuel consumption average
of 15.59

Figure 7 shows the engine speed in function of the gear shifting strategy used and the way that hybrid strategy operates
by changing the default gear shifting speed according to the vehicle power demand.

Figure 7 - Gear shifting tactics utilized

Figure 8 shows the gear shifting strategy changes managed by the implemented algorithm. In the high acceleration
stretches, is used the maximum power strategy to the vehicle reach the standard velocity profile the required
performance. Already in low power demand situations fuel economy tactic is utilized. The intermediary strategies are used
in moderate acceleration stretches. The maximum torque strategy is used in shorts high acceleration stretches where the
vehicle reachs the desired speed before the engine reaches maximum power strategy high speeds.

Figure 8 - Engine speed in function of the gear shifting tactic

In the Figure 9 and 10 can be observed the differences in the gear profiles used by the fuel economy and hybrid
strategies. While in the simulations with fixed gear shifting strategies gear profiles are uniform due to upshifting and
downshifting occur at specifics speeds stipulated by the gear shifting tactic used. The hybrid strategy possibility the
downshift process even in acceleration stretches to increase the vehicle wheels torque, increasing the vehicle's
performance and consequently the fuel consumption.

Figure 9 - Fuel economy tactic gear profiles

 Figure 10 - Hybrid strategy gear profiles

CONCLUSION

In this paper gear shifting strategy influence in the vehicle longitudinal calculation dynamics was evaluated. The results
were obtained by co-simulations performed in the multibody dynamics analysis Adams TM with Simulink/MatlabTM.

The calculation methodology was described in the book Fundamentals of Vehicle Dynamics, considering the vehicle
geometry and powertrain inertias of a Compact Hatchback, equipped with 1.0L engine and also a gearbox with five ratios.

After the first simulation, it was observed that the vehicle performance was lower than expected in some stretches where
engine maximum power was required, due to high accelerations imposed by the US06 speed profile.

With the purpose to extend the permanence time using the greatest reduction gear ratios (generating a higher available
torque at the vehicle wheel) gear shifting strategies at higher engine speeds have been implemented. It was verified a
gain in the vehicle performance and consequently an increase of fuel consumption due to new gear shifting strategies
make the engine run in lower efficiency regions.

Aiming to keep the maximum power strategy performance, and reducing the fuel consumption, was developed a hybrid
strategy that combines the previously simulated gear shifting tactics applying them as required.

The hybrid strategy resulted in a similar correlation of the maximum power tactic, but with lower fuel consumption close to
the results found using the intermediary strategies. The hybrid strategy improves a better representation of the driver
behavior in function of the required performance, which is provided by the US06 standard speeds profile.

Finally is concluded that the knowledge of the gear shifting strategies is crucial in the vehicle dynamics calculation,
because these influence the vehicle performance and consequently the fuel consumption.

ACKNOWLEDGMENTS

The authors thank CNPq, CPFL, and ANEEL for financial support.
REFERENCES

1. Brezina, Tomas, Zdenek Hadas, and Jan Vetiska. "Using of Co-simulation ADAMS-SIMULINK for development of
mechatronic systems." MECHATRONIKA, 2011 14th International Symposium. IEEE, 2011.
2. Casavola, Alessandro, Giovanni Prodi, and Giuseppe Rocca. "Efficient gear shifting strategies for green driving
policies." American Control Conference (ACC), 2010. IEEE, 2010.
3. Corrêa, F.C., Silva, L.C. and Dedini, F.G., 2011. “Fuzzy control for hybrid vehicle”. 21st Brazilian Congress of
Mechanical Engineering, Vol. 1, p. 9.
4. Ehsani, Mehrdad, Yimin Gao, and Ali Emadi. Modern electric, hybrid electric, and fuel cell vehicles: fundamentals,
theory, and design. CRC press, 2009.
5. Gao, Bingzhao, et al. "Observer-based clutch disengagement control during gear shift process of automated
manual transmission."Vehicle System Dynamics 49.5 (2011): 685-701.
6. Gillespie, T. D., 1992. Fundamentals of vehicle dynamics. Society of Automotive Engineers, Warrendale, 1st
edition.
7. GM, 2013, "Manual do proprietário Chevrolet Celta 2013,". General Motors do Brasil Ltda.
8. Guan, Tianyi, and Christian W. Frey. "Fuel efficiency driver assistance system for manufacturer independent
solutions." Intelligent Transportation Systems (ITSC), 2012 15th International IEEE Conference on. IEEE, 2012.
9. Milliken, W.F., Milliken, D.L. and Metz, L.D., 1995. Race car vehicle dynamics, Vol. 146. Sae International
Warrendale, PA.
10. Kang, Mingxin, et al. "Coordinated vehicle traction control based on engine torque and brake pressure under
complicated road conditions." Vehicle System Dynamics 50.9 (2012): 1473-1494.
11. Oliveira, E.C., 2005. Modelagem e simulação de veículos elétricos e híbridos. Master’s thesis, Escola Politécnica
da Universidade de São Paulo.
12. Orfila, Olivier, Guillaume Saint Pierre, and Cindie Andrieu. "Gear Shifting Behavior Model for Ecodriving
Simulations Based on Experimental Data." Procedia-Social and Behavioral Sciences 54 (2012): 341-348.
13. Reimpell, J. and Stoll, H., 1996. The automotive chassis: engineering principles. Society of Automotive Engineers
Warrendale„ PA, USA.
14. Santiciolli, F. M., Costa, E. S., Eckert, J. J., Dionísio, H. J. and Dedini, F. G., 2013,"Gear Shifting Optimization
Strategy for Brazilian Vehicles and Traffic" Proceedings of the 22th International Congress of Mechanical
Engineering, Ribeirão Preto, Brazil.
15. Serrao, L., a.C.Z., Guezennec, Y. and Rizzoni, G., 2005. “An aging model of ni-mh batteries for hybrid electric
vehicles”. IEEE Vehicle Power and Propulsion Conference (VPP05), Vol. 1, pp. 78–85.
16. Singh, J., et al. "Selection of Gear Ratio for Smooth Gear Shifting." Analysis 1 (2012): 1995.
17. Vagg, C., Brace, C. J., Wijetunge, R., Akehurst, S., and Ash, L., 2012. "Development of a new method to assess
fuel saving using gear shifting indicators." In Proceedings of the Institution of Mechanical Engineers, Part D:
Journal of Automobile Engineering 226.12: 1630-1639.
18. Wong, Jo Yung. Theory of ground vehicles. Wiley. com, 2001.
19. Xi, Lu, Xu Xiangyang, and Liu Yanfang. "Simulation of Gear-shift Algorithm for Automatic Transmission Based on
MATLAB." Software Engineering, 2009. WCSE’09. WRI World Congress on. Vol. 2. IEEE, 2009.
20. Yin, Xiaofeng, Dianlun Xue, and Yun Cai. "Application of time-optimal strategy and fuzzy logic to the engine
speed control during the gear-shifting process of AMT." Fuzzy Systems and Knowledge Discovery, 2007. FSKD
2007. Fourth International Conference on. Vol. 4. IEEE, 2007.
21. Yu, Wei-Bo, Ding-Xuan Zhao, and Chen Nimg. "Study on fuzzy gearshift tactics in automatic gearshift control
system of technical vehicles." IEEE Industial Electronics Society. Annual conference. 2004.

CONTACT

javorski@fem.unicamp.br, fabio@fem.unicamp.br, eduardo.costa@fem.unicamp.br, hjd_92@hotmail.com, dedini@fem.

unicamp.br

View publication stats