Skip to main content

LUP Student Papers

LUND UNIVERSITY LIBRARIES

Optimal Control of Multi-Speed Transmissions in Electric Vehicles

Abrahamsson, Fredrik LU (2020) In CODEN:LUTEDX/TEIE EIEM01 20192
Division for Industrial Electrical Engineering and Automation
Abstract
The increasing electrification in the automotive industry places new demands on the companies and on the technology. BorgWarner Inc. is investigating the possible gains of implementing a multi-speed transmission
in an electric vehicle. They are developing a new double clutch with low losses and low torque interrupts.
The clutch is fitted in a gearbox which allows the electric powertrain to move the operating point to a more efficient point.
Previous simulations showed promising results for the multi-speed transmission but the simulation model required improvements before any conclusions could be made. The simulation model is improved with a
new battery model, new clutch model, new transmission model and rotational inertia. The model... (More)
The increasing electrification in the automotive industry places new demands on the companies and on the technology. BorgWarner Inc. is investigating the possible gains of implementing a multi-speed transmission
in an electric vehicle. They are developing a new double clutch with low losses and low torque interrupts.
The clutch is fitted in a gearbox which allows the electric powertrain to move the operating point to a more efficient point.
Previous simulations showed promising results for the multi-speed transmission but the simulation model required improvements before any conclusions could be made. The simulation model is improved with a
new battery model, new clutch model, new transmission model and rotational inertia. The model can no longer be controlled by a controller which changes gear based on the efficiency of the electric machine. By
calculating the optimal control signal the optimal energy consumption is obtained.
The backtracking method dynamic programming is used to obtain the optimal control signal for the vehicle.
The optimal controller is then used to optimize the gear ratios in the transmission as well as a benchmark for the online controller. The online controller can be installed in a real vehicle because it uses data from the
past. The performance of a simple rulebased controller which just considers the speed of the vehicle is better than expected and chosen as controller. By optimizing the shifting point, the consumption for the vehicle with the online controller is less than 1% higher than the optimal. To improve the controller acceleration is taken into consideration but the improvements are not sucient to justify the increased complexity.
Sensitivity analyses of the vehicle show that a third gear lowers the consumption but not enough to give grounds for an implementation. The analysis is made on several drive cycles which concludes that the shifting point can be optimized for a specic driving style but the gain is not remarkable. (Less)
Popular Abstract
The climate crisis is today a hot topic and the European Commission is aiming to lower the green house gas emissions by 60% by year 2050. To succeed with the task the automotive industry is continuously developing more ecient and lighter vehicles. Electric vehicles are pointed out as a possible solution to lower the emissions from the transportation sector.
Electric vehicles use electricity instead of fossil fuel as energy source. The benefits with electricity are the high efficiency in the powertrain, the low noise and almost zero emissions from the tailpipe. An issue with
an electric vehicle are the batteries. A fuel tank with fossil fuel contains more energy than a battery of the same weight. By increasing the eciency of the... (More)
The climate crisis is today a hot topic and the European Commission is aiming to lower the green house gas emissions by 60% by year 2050. To succeed with the task the automotive industry is continuously developing more ecient and lighter vehicles. Electric vehicles are pointed out as a possible solution to lower the emissions from the transportation sector.
Electric vehicles use electricity instead of fossil fuel as energy source. The benefits with electricity are the high efficiency in the powertrain, the low noise and almost zero emissions from the tailpipe. An issue with
an electric vehicle are the batteries. A fuel tank with fossil fuel contains more energy than a battery of the same weight. By increasing the eciency of the powertrain in a vehicle the range of the vehicle can be
extended or the battery could be smaller.
Most electric vehicles today are using only one gear. The reason for this is that an electric motor has constant power over a wide speed range. Thus it does not need many gears unlike a combustion engine.
Lately high-end cars like the Porsche Taycan and electric trucks like the Volvo FL Electric installed a twospeed transmission to the powertrain. For the Taycan it means that a high starting torque and a high top speed can be achieved. For the Volvo it means that it has the torque to drive over curbs despite its high weight and still reach 90km/h on the highway.
At BorgWarner the question is if a multi-speed transmission can be beneficial for a C-segment passenger car as well. This thesis investigates the potential benefits a multi-speed transmission has in a small five door
SUV. Preliminary simulation showed a potential reduction in energy consumption by 2.4% when a two speed transmission was implemented. The simulation model needed further improvements for any conclusions to be drawn. By implementing a new battery model, new transmission model, include the rotating inertias and implement slip losses in the clutch a more realistic vehicle model is achieved.
To be able to ensure the optimal performance of the vehicle an optimal control algorithm based on dynamic programming is developed. Dynamic programming nds the best possible way to change gears which consumes the least possible amount of energy. The optimal control algorithm is an effective tool to analyse the vehicles performance. The gear ratio between gear one and gear two is optimized by running the dynamic programming script for a wide range of ratio combinations to nd the optimal. The consumption of the vehicle with the new gear ratios is 0.1426kWh/km which can be compared to the 0.1486kWh/km for the single-speed transmission which is a decrease in energy consumption by 4%. The analysis is extended to a three-speed transmission to evaluate the eciency gain with one extra gear. The energy consumption for a three-speed transmission is lower but not enough to cover the extra cost for extra components and development.
The optimal controller can not be used in a real vehicle because the dynamic programming algorithm require knowledge of the entire drive cycle before is starts. This is not possible in a vehicle due to unexpected event
such as trac and changes in weather etc. An online controller is developed to create an understanding of how close to the optimal a real controller can perform. The development of the controller started basic with a speed dependent controller just using the speed as input. The controller shifts to second gear when the speed is above a threshold and down when it is bellow it. The performance of the controller is better than expected. The shifting point are optimized to shift at 51.1 km/h which results in an energy consumption just 0.4% higher than the optimal and 3.7% lower than the single speed version. (Less)
Please use this url to cite or link to this publication:
author
Abrahamsson, Fredrik LU
supervisor
organization
course
EIEM01 20192
year
type
H3 - Professional qualifications (4 Years - )
subject
keywords
Optimal Control, Electric vehicle, Multi-Speed Transmission, Matlab
publication/series
CODEN:LUTEDX/TEIE
report number
5428
language
English
id
9004577
date added to LUP
2026-09-28 11:46:09
date last changed
2026-09-28 11:46:09
@misc{9004577,
  abstract     = {{The increasing electrification in the automotive industry places new demands on the companies and on the technology. BorgWarner Inc. is investigating the possible gains of implementing a multi-speed transmission
in an electric vehicle. They are developing a new double clutch with low losses and low torque interrupts.
The clutch is fitted in a gearbox which allows the electric powertrain to move the operating point to a more efficient point.
Previous simulations showed promising results for the multi-speed transmission but the simulation model required improvements before any conclusions could be made. The simulation model is improved with a
new battery model, new clutch model, new transmission model and rotational inertia. The model can no longer be controlled by a controller which changes gear based on the efficiency of the electric machine. By
calculating the optimal control signal the optimal energy consumption is obtained.
The backtracking method dynamic programming is used to obtain the optimal control signal for the vehicle.
The optimal controller is then used to optimize the gear ratios in the transmission as well as a benchmark for the online controller. The online controller can be installed in a real vehicle because it uses data from the
past. The performance of a simple rulebased controller which just considers the speed of the vehicle is better than expected and chosen as controller. By optimizing the shifting point, the consumption for the vehicle with the online controller is less than 1% higher than the optimal. To improve the controller acceleration is taken into consideration but the improvements are not sucient to justify the increased complexity.
Sensitivity analyses of the vehicle show that a third gear lowers the consumption but not enough to give grounds for an implementation. The analysis is made on several drive cycles which concludes that the shifting point can be optimized for a specic driving style but the gain is not remarkable.}},
  author       = {{Abrahamsson, Fredrik}},
  language     = {{eng}},
  note         = {{Student Paper}},
  series       = {{CODEN:LUTEDX/TEIE}},
  title        = {{Optimal Control of Multi-Speed Transmissions in Electric Vehicles}},
  year         = {{2020}},
}