N. Ligterink, R. Cuelenaere RCG, 8 September 2014 N. Ligterink, R. Cuelenaere
Random Cycle Generator chronology RCG, 8 September 2014 Random Cycle Generator chronology alpha version specifications from the Commission alpha version distributed feedback: minor problems with cycle generation and signal quality further development derived from RDE-PEMS development: fixed selection from the available options in the alpha version cycle requirements: alignment with RDE trip requirements shift from “time-based” to “distance-based” Validation: drivability representativeness Add-ons…. current status follow-ups
RCG components distance distribution urban/rural/ motorway RCG, 8 September 2014 RCG components distance distribution urban/rural/ motorway vehicle information WLTP database probabilities of driving behaviour random cycle generator constraints and boundary conditions cycle validation driving cycles check testing (outside project scope) method validation
Methodology Based on data from WLTP database (European driving) RCG, 8 September 2014 Methodology Based on data from WLTP database (European driving) Markov chain approach: probabilities of changing acceleration and velocity band stochastic approach two time scales: seconds (for acceleration) minutes (for average velocity) Restrictions from vehicle and test limitations (i.e. maximal power) Each cycle to cover urban, rural, and motorway driving: derived from the RDE-PEMS trip requirements transitions between the ten velocity bands:
Markov matrix current state region of high probability start braking RCG, 8 September 2014 current state when decelerating remain decelerating input region of high probability start braking deceleration driving at a constant velocity low probability next state P selection start accelerating when accelerating remain accelerating acceleration low probability Markov matrix deceleration acceleration typical periods (2-10 seconds) of accelerations in a real-world driving
Random cycles? Different speed profiles (visual) Statistical analyses RCG, 8 September 2014 Random cycles? Different speed profiles (visual) Statistical analyses
Statistical analyses of generated cycles average velocity over cycle RCG, 8 September 2014 average velocity over cycle velocity distribution in road types urban rural motorway distance between stops Statistical analyses of generated cycles
Open issues RDE-PEMS trip requirements are not cycle requirements: RCG, 8 September 2014 Open issues RDE-PEMS trip requirements are not cycle requirements: >120 minutes is too long on a chassis dynamometer short times, with equal distance, gives a very short motorway part gear shift strategies: free choice/OEM/GSI Heinz Steven tool (WLTP) fixed velocities (NEDC) fixed velocities in power-to-mass classes (CADC) delayed gear shifts during acceleration (TNO-approach) driver assistance: difficult to “follow a line” on the screen: driving simulation?
Maximum cycle duration? Change distance per road type? RCG, 8 September 2014 Reflections? Maximum cycle duration? Change distance per road type?
Open issues Preferences? RCG, 8 September 2014 Open issues RDE-PEMS trip requirements are not cycle requirements: >120 minutes is too long on a chassis dynamometer short times, with equal distance, gives a very short motorway part gear shift strategies: free choice/OEM/GSI Heinz Steven tool (WLTP) fixed velocities (NEDC) fixed velocities in power-to-mass classes (CADC) delayed gear shifts during acceleration (TNO-approach) driver assistance: difficult to “follow a line” on the screen: driving simulation? Preferences?
Next phases Beta-version and draft report asap RCG, 8 September 2014 Next phases Beta-version and draft report asap Concept for Validation: TUG end ‘14 Legal text early ‘15 Validation tests (outside scope project) early ‘15
RCG, 8 September 2014 Back up slide
Alpha-version - Input to the random cycle generator RCG, 8 September 2014 Alpha-version - Input to the random cycle generator examples: Prated = 82 kW M = 1412 kg F0 = 153 N F1 = 0.54 N/[km/h] F2 = 0.056 N/[km/h]2 physical vehicle limitations based on type-approval data: test mass rated power road-load values chassis dynamometer restrictions (optional) avoiding tyre slip, by limiting maximal deceleration cycle limitations: (optional) total duration (# seconds) percentages, and order, of urban/rural/motorway(/ex) amin > - 2 m/s2 Ncycle = 1800 sec urban rural motorway urban = 27% [distance] rural = 31% [distance]