Range Rover Evoque as a development platform
Flanders Make offers infrastructure for testing and validating products and production systems. As part of these activities, we have adapted two Range Rover Evoques. We developed a hybrid version and an all-electric version that we use for all kinds of research and testing. Both by our own researchers and by our customers.
Our hybrid Evoque is equipped with a P4 hybrid powertrain where the front wheels are driven by the original combustion engine and corresponding transmission. Every rear wheel is driven by an in-wheel electric motor. Finally, this vehicle is also equipped with active suspension on the rear wheels.
Our electric Evoque is equipped with 4 electric motors (one for each wheel) and a hydraulic brake-and-steer-by-wire system so that the vehicle can drive completely autonomously. The table below shows the main specifications of both vehicles.
|Electric Evoque||Hybrid Evoque|
(4x switched reluctance motor)
|Power||100 kW per wheel||Front: 177 kW
Rear: 2 x 65 kW
|Transmission||1/10 per wheel||Front: 9-speed automatic transmission
Rear: No transmission
|Battery (voltage, energy)||600 V, 9 kWh||350 V, 13 kWh|
|Battery cell technology||LI-Ion, LTO||Li-Ion, NCA|
|Weight||2250 kg||2170 kg|
|By-wire functionality||4 x traction
4 x hydraulic brakes and steering
|Traction: both the combustion engine at the front and both electric motors at the rear.|
|Control||Rapid prototyping system||Rapid prototyping system|
|Suspension||Front/rear: passive||Front: passive
Rear: active (1 kN peak force at each wheel)
Vehicle dynamics controllers
An important part of our innovations revolves around vehicle dynamics controllers. A few examples:
- A controller for an electric ABS system using the electric motor per wheel instead of the conventional hydraulic system;
- A controller that allows electric cars with one electric motor per wheel to drive more efficiently in all conditions;
- A fault-tolerant lateral controller that ensures that if there is a fault in one of the drives (if a motor fails) the vehicle will still follow the desired path;
- Development of a controller that leads to improved driving comfort. We achieved this by optimising the control of the active suspension on the rear wheels. This applies to both road bumps and the dynamic properties of the vehicle. A short summary of this can be found in the following video:
Report on winter tests in Sweden
We recently had the opportunity to take our hybrid Evoque to the next TRL level by participating in DANA's winter tests. The Swedish surface of snow and ice provides little grip. As a result, slippage occurs at lower speeds and the forces on the powertrain are much lower. This allows us to develop the controllers more safely and easily so as to achieve the desired vehicle behaviour.
To be able to operate in the freezing Swedish climate, we made our vehicle winter-proof with, amongst other things, adapted cooling, tyres, etc. In addition, we have adapted and optimised previously developed model-based controllers, namely the torque vectoring algorithm, by performing tests in various driving scenarios.
The different driving scenarios of the tests include:
- Driving in circles with a constant radius;
- Driving on handling tracks;
- Performing the well-known Moose test.
By gradually increasing the speed while driving in circles with a constant radius, we can determine the understeer coefficient of the vehicle. This value indicates how strongly the vehicle tends to either (a) go into unstable oversteer (drifting), or (b) go into a more stable understeer, where the vehicle no longer responds to steering.
On the handling track and during the Moose tests, we evaluate the dynamic driving behaviour of the vehicle. Here, we look not only at longitudinal and lateral accelerations, but also at vertical accelerations, roll speed and pitch speed. In this way, we also map out the comfort level of the passengers.
Finally, the Moose test is an important test to improve the safety of the car. This test simulates swerving for a moose on the road without losing control. Here, too, our systems such as torque vectoring and active suspension ensure better performances. The figure below shows three moose tests with different vehicle settings: with Active Suspension (AS) and Torque Vectoring (TV), with TV only or with both systems switched off. This figure shows that for a similar lateral acceleration and yaw rate (rotational speed of vehicle about its vertical axis) the vehicle experiences less roll. This increases comfort and improves road holding.
During these tests, we use a dSPACE rapid prototyping platform to enable us to easily log data and make adjustments to the control software. This allows us to analyse the data in both the time and frequency domain immediately after the tests and thus improve our control software before testing the next iteration.
After three weeks of testing and optimising the various control parameters, our vehicle performs better and in a more predictable way in winter conditions compared to a standard vehicle without active control. In addition, due to the active control of the rear suspension, the passenger comfort during the test drives also increased.
The next Evoque steps
The next steps we want to take with our vehicles are:
- More accurate vehicle state estimators: the aim here is to determine the vehicle status even more accurately and with fewer sensors and to use these data in the control algorithms. In concrete terms, this will lead to better algorithms and reduced costs as fewer sensors will be needed;
- Development of a scalable, modular software framework that can be used for a wide range of vehicles. This framework must also be compatible with more complex adaptive hardware and support the challenges towards autonomous driving.