E-VECTOORC, A Model of Cooperation

Transcription

1 E-VECTOORC, A Model of Cooperation Phil Barber Automotive Electronic Systems Innovation Network (AESIN) UK Automotive Electronics Conference October 2014

2 Contents Electric Vehicle Control of Individual Wheel Torque for On- and Off-Road Conditions Who What When (Details) Dissemination

3 Who Project Coordinator: Aldo Sorniotti (University of Surrey, UK)

4 Who Skoda JLR

5 Who Inverto is a technology company that designs and develops power electronics products. Flanders Drive are based close to the Lommel test track and offer management support in setting up research projects and access to advanced research infrastructure and test tracks. Punch Powertrain produce 400K transmission units a year and are researching into CVTs and Hybrid powertrains.

6 Who Fundacion CIDAUT is a Spanish non-profit research and development centre in transport and energy. The Instituto Tecnológico de Aragón (ITA) is a public non-profit research centre, established in 1984 by the regional government of Aragon TRW Koblenz

7 Who Ilmenau University of Technology University of Surrey VIF Kompetenzzentrum - Das virtuelle Fahrzeug, Forschungsgesellschaft mbh. an independent, international platform for research and development in the automotive industry.

8 What (for video see

9 What The project was to investigate: How performance can be improved: Torque distribution Off road capability Vehicle dynamics Enhanced safety on the limit of adhesion How economy can be improved: Reduced parasitic losses Regenerative braking Investigation into opportunities for a detachable / trailer range extender This was achieved by: Delivery of a working 4eV prototype Delivery of a brake by wire vehicle Extensive CAE and Simulation studies [M1 Report]

10 When 10/10/12 Technical review, Brussels 19/3/13 18 mth Review at Skoda 13/5/13 Dissemination Event,Graz 4WD Vehicle (Half shaft issues) E N D BBW HIL tests, Ilmenau Motor tests, Inverto 2WD Vehicle Tests Lommel Torque vectoring Vehicle Tests 4WD Vehicle build 600V Battery Vehicle Tests NEDC Dynamics EMC Dissemination Day Range Extender trailer

11 Details 4WD, 4 x Electric SR, inboard motor Evoque 600V, 9KWh battery pack TRW Brake by wire system + HIL rig at Ilmenau University Brake Blending and eabs Torque apportionment algorithms for understeer gradient compensation. Active (torsional) vibration control (electric anti shunt) Range extender trailer

12 4WD Electric Evoque Details in Deliverable D7.2

13 Electric Motor Characteristics 600V Peak (30s) 200Nm, 75kW Nominal (continuous) 80Nm, 35kW Maximum speed min-1 Motor dimensions (LxD) 215x265 mm Motor weight 50 kg Motor inertia (without gearbox) kgmm² Inverter dimensions (WxHxD) 495x155x282 mm Inverter weight 16.2 kg Liquid cooled 15l/min, 55 C max inlet Operating temperature C HIL rig based efficiency maps of the motors in both regen (peak = 91%) and motoring (peak = 89) have been created by Inverto Torque estimators have been generated: one using using efficiency maps, one using current and rotor position A clear trade between torque ripple and efficiency was demonstrated extending the phase overlap reduces torque ripple, but adversely impacts efficiency.

14 Battery Characteristics Pack nominal voltage 600V Pack maximum voltage V Pack minimum voltage 414 V Peak current charge and discharge 400A Nominal current (continuous) charge and 200A discharge Peak power charge and discharge 310 kw Nominal power charge and discharge 121 kw Cell capacity 15Ah Energy 9.1kWh Mass Estimated 300kg (225 kg for current 400V pack) Operating temperature 0 55 C

15 Brake System Controllers of both foundation and regenerative braking through TRW s BBW and emotor control for ABS and stability control. Brake blending control algorithms Braking controller integrated with Carmaker, HIL rigs and the rest of the E-VECTOORC controller. EXPERIMENTAL INVESTIGATION OF BRAKING DYNAMICS OF ELECTRIC VEHICLE Ilmenau University of Technology, Germany, TRW Automotive, Germany

16 Control System Active Vibration Controller Brakes Blending Controller ABS and eabs Slip Control algorithm Figure Wheel speed and motor speed (rad/s) results for AVC ON and AVC OFF. 30kph and low torque conditions.

17 Active Vibration Control Synthetic drivetrain (gearbox / halfshaft) damper Electric version of anti-shunt. Consequent of 10:1 gearbox inboard motor. Figure Wheel speed and motor speed (rad/s) results for AVC ON and AVC OFF. 60kph and braking conditions. Figure Wheel speed and motor speed (rad/s) results for AVC ON and AVC OFF. 30kph and low torque conditions.

18 ABS and eabs Gains scheduled with speed Speed over ground state estimation, filters to establish slip Closed loop frequency domain analysis Stable slip target achieved. Figure 3.1 ABS and eabs control structure 0.05 Front left wheel Slip [/] Unfiltered Filter Kalman Time [s] Figure 3.5 Slip controller frequency response Figure 3.9 High bandwidth (10Hz) speed signal with increased proportional gain. Figure 3.3 Filtering of the speed signals and the effect on slip estimate.

19 Brake Blending Controller The brake blending algorithm is working correctly, generating a correct brake allocation and keeping the defined limits. Deceleration signal has no peaks or abrupt changes when there are transitions between both brake sources. Deceleration has slow variations along the test, whilst a constant value was expected. General tendency is to have a reduction when regenerative is increasing or maintained; when only friction is applied, higher deceleration is obtained. When only friction is considered, deceleration results are constant. Similar test should be done with regenerative brake only, to analyze the deceleration reduction obtained in the brake blending tests. Oscillations are obtained in longitudinal acceleration when the vehicle completely stops, with low frequency (possibly pitch) resonance. Pressure is released approx 1 second later, and no effect on acceleration is noticed. FL Torque Friction Motor Time (sec) 2 1 Vehicle accelerations Time (sec) Figure % speed limit case. Brake blending results in N m at the front-left motor

20 Vehicle Dynamics (Slip Control) Comparing vehicle modes which present the same reference yaw rate (i.e. baseline, normal mode and ESP mode), it can be seen that the vehicle in normal mode presents the smallest value of yaw rate overshoot and sideslip angle peak. When the sport mode tuning is chosen, the values of yaw rate overshoot and sideslip angle peak are considerably lower than the respective values measured for the baseline vehicle, although the sport mode presents more aggressive behaviour than the baseline vehicle itself. Looking at the plot of the longitudinal vehicle speed during a step steer, the comparison between the vehicle modes which present the same reference yaw rate (baseline, normal mode, ESP mode) shows that the reduction of the vehicle speed for the normal mode is lower than the one evaluated for the baseline vehicle because of the increased induced drag generated for the baseline vehicle. This is consequent to the larger values of lateral slip. The vehicle in ESP mode decelerates more with respect to the vehicle in normal mode because of the braking action performed by the actuation of the friction brakes for the understeer/oversteer correction during the step steer manoeuvre. normal mode 100 sport mode baseline 80 δ [deg] Figure Skid pad tests carried out at radius R = 60 m for the 2WD version of the vehicle demonstrator. Baseline vehicle (dot-dashed line), torque-vectoring vehicle in normal mode (solid line) and torque-vectoring vehicle in sport mode (dashed line a y [m/s 2 ]

21 Trailer Range Extender Requirements 20kW installed, 4-600V delivered To support 2hrs continuous usage of vehicle It was resolved by the consortia that this range extender market is very fast moving. Direct competition in this market was not of primary interest to the development of the project aims, however a new concept of a demountable range extender was to be investigated. This would both support the track testing and investigate the utility and acceptability of this concept for inter-city hybrid mobility. The target mass of the developed trailer was to be close to the 750Kg threshold for trailers where there is a breakpoint in the driver licencing requirements. The trailer had to be aesthetically appealing to the market.

22 Trailer Range Extender Requirements 20kW installed, 4-600V delivered To support 2hrs continuous usage of vehicle Trailer Developed: A trailer, styled to match the RR Evoque was developed using off the shelf hardware, including a standard diesel engine and power electronics. The styling was coordinated with the E-VECTOORC vehicle. Weight: 800Kg Generator output 400Vac, 3phase, 22KVA (19.8KW), Charging Output voltage max800vdc Charging Output current max 25Amps + Single phase socket 230Vac, 16amp Control, Starting and charging 12Vdc 12V Battery capacity 95Ah, 850 CCA

23 Public Dissemination Event 28 th August, Heritage Centre Session 4: Vehicle longitudinal dynamics and control E-VECTOORC Dissemination Event Agenda (Draft) (v.2) Heritage Centre, Gaydon 28th August 2014 Organisers Phil Barber, Nigel Clarke (Jaguar Land Rover) Valentin Ivanov (Ilmenau University of Technology) Aldo Sorniotti (University of Surrey) 09:30 Welcome / timetable of the event (Mike Richardson) 09:45 E-VECTOORC project and objectives overview (University of Surrey) Session 1: Demonstrator vehicle construction 10:00 Vehicle requirements, architecture and build (Flanders Drive) 10:30 Switched reluctance motors (Inverto) 16:00 Friction brake system (by-wire) implemented on the vehicle demonstrator (tba, TRW; Dzmitry Savitski, Ilmenau University of Technology) 16:30 Torsional vibration control of on-board electric drivetrains (ITA) 17:00 Regenerative braking and longitudinal slip control using electric machines (Dzmitry Savitski; Ilmenau University of Technology; tba, ITA) Session 5: Closing remarks and open discussion of other projects related to E-VECTOORC 17:30 Other projects related to E-VECTOORC (Aldo Sorniotti, University of Surrey) Session 2: Safety and reliability aspects of fully electric vehicles 11:00 EMC testing for fully electric vehicles (ITA) 11:30 Functional safety for fully electric vehicles (Flanders Drive, CIDAUT) Session 3 : Direct yaw moment controllers 14:00 Objectives of the E-VECTOORC yaw moment controller and analyses based on the quasi-static model (University of Surrey) 14:30 Comparison of yaw moment controller formulations (University of Surrey, ViF) 15:00 Experimental results and benchmarking (University of Surrey, Jaguar Land Rover and SKODA Auto)

24 References: guarlandrover.com/evectoorc/

25 Pictures, Video

26 Next: Feed of results into other research and delivery programs Further research inside JLR... Further EC research, this time JLR as 'Assisting Partner'

27 Conclusions: Strong consortia (mix) led by capable and proactive project manager reporting to Project Officer at EC. Vehicle was pivotal to motivation and to validation.. It had to work. Physically remote, but good team building events due to nature of travel (extra mural events). Expense of travel. Time out of office (or away from the family) travelling. Budget management overhead not to be underestimated, financial support by EC (and ratio rules). Deadlines imposed by EC to meet and deliver by (e.g. dissemination activities). Significant commitment is required. 'Why contribute', the secondary agendas are sometimes the most important. Relationship building for the consultative partners or potential start up suppliers.

28 Questions?