Fakultät Verkehrswissenschaften „Friedrich List“ - Institut für Automobiltechnik Dresden – IAD - Lehrstuhl Kraftfahrzeugtechnik Faculty of Transportation and Traffic Sciences „Friedrich List“ - Institute for Automotive Technologies Dresden - Chair of Automotive Engineering "auto.mobile-driving simulator" - a new immersion IPG Apply & Innovate 2016 Tüschen, T .; Prokop, G.
Agenda 1 Motivation 2 State-of-the-art driving simulator 3 auto.mobile-driving simulator 4 Prediction 21.09.16 – Karlsruhe IPG Apply & Innovate ‘16 – Thomas Tüschen 2
1. Motivation Evaluation of Advanced Driver Assistance Systems Passive safety Active safety / ADAS Source: Euro NCAP Source: Euro NCAP Ratable by means of the cars Ratable only on the basis of the interaction of the technical properties: driver and the car under realistic environmental impact (e.g. traffic): occupant values probability of an accident intrusion values potential accident severity Mostly ex-post evaluation possible A-priori evaluation possible based on analysis of sufficient numerous accident due to crash-tests and simulation statistics 21.09.16 – Karlsruhe IPG Apply & Innovate ‘16 – Thomas Tüschen 3
1. Motivation Use of driving simulator Driving Simulator Driving scenarios Dynamics Source: Verlag Heinrich Vogel Source: DLR Source: Toyota Source: Daimler Advanced Driver Assistance Systems & Automated Driving Driving Dynamics (also part of ADAS) Source: Bosch Source: BMW AG high variety of driving scenarios high (horizontal-) dynamic necessary highly immersive driving simulators necessary unscaled motion perception Critical exemplary scenario: Critical exemplary scenario: low frequencies, low acceleration high frequencies, high acceleration 21.09.16 – Karlsruhe IPG Apply & Innovate ‘16 – Thomas Tüschen 4
2. State-of-the-art driving simulator Requirements to state of the art driving simulator Statement 1 – Use of Tilt-Coordination * only at low frequencies possible Tilting rate threshold ( 𝜒 𝑚𝑗𝑛𝑗𝑢 ): 5 °/s max. acceleration gradient: 9,81 m/s²* sin(5 °/s ) = 0,85 m/s² /s Real translational motion mandatory! Source: Verlag Heinrich Vogel Statement 2 – Free yaw motion requires a double sled system F lat F lat 𝜔 𝜔 Statement 3 – Motion without frequency gaps requires large motion space Source: DLR Required motion space for an optimal unscaled motion simulation: ±161m Source: Daimler Source: Dissertation Betz Statement 1+2+3 System with large x,y-motion space is necessary! *Tilt-Coordination: The tilting of the subject leads to a transformation of the gravity force vector in order to simulate a horizontal acceleration. Sustained Source: Toyota accelerations are simulated by tilting the simulator. ( 𝑏 𝑢𝑗𝑚𝑢 = 9.81 𝑛 𝑡 2 ∗ sin 𝜒 𝑢𝑗𝑚𝑢 ) 21.09.16 – Karlsruhe IPG Apply & Innovate ‘16 – Thomas Tüschen 5
2. State-of-the-art driving simulator Requirements to state of the art driving simulator Sledge Source: Verlag Heinrich Vogel Hexapod Force and Energy calculation: Source: DLR m Hexapod = 4t; m Sledge = 16t; a = 9,81 m/s²; l Sledge = 15m; v max ≈ 12m/s² F Hexapod = 4t*9,81m/s² ≈ 39kN F Sledge = ( 4t + 16t)*9,81m/s² ≈ 196kN Source: Daimler E Hexapod = 0,5* 4t*(12 m/s)² ≈ 288kJ E Sledge = 0,5*( 4t + 16t)*(12m/s)² ≈ 1440kJ Expansion from a 1- to a 2-sledge system, with a sledge length of only 15m, would increase the force and energy demand by factor 5. Source: Toyota 21.09.16 – Karlsruhe IPG Apply & Innovate ‘16 – Thomas Tüschen 6
3. auto.mobile-driving simulator Concept Details Modifiable cockpit Visualization screen concept Motion platform Accumulator – Tripod (3 DOF) Ring bearing Dual suspension (1 DOF) 4 Wheel pairs Main structure with 8 electric (3 DOF) motor Source: AMST-Systemtechnik GmbH DOF – Degree of Freedom 21.09.16 – Karlsruhe IPG Apply & Innovate ‘16 – Thomas Tüschen 7
3. auto.mobile-driving simulator Mobility Concept Source: AMST-Systemtechnik GmbH Driving simulator center General driving area Source: AMST-Systemtechnik GmbH 21.09.16 – Karlsruhe IPG Apply & Innovate ‘16 – Thomas Tüschen 8
3. auto.mobile-driving simulator Problems of a wheel-based system Tire as a non holonomic constraint Motion of the simulated/real car ≠ Motion of the 𝜄 t =0 driving simulator (e.g. car is driving and simulator ψ t =const. stands still) Active build-up of lateral forces only possible with a rolling speed ≠ 0 z t ψ t Tire must never stand still φ t θ t x t y t F y =0 Rotational speed depending phase delay Tire has a relaxation length 𝜄 t ≠0 ψ t =const. No sudden build-up of lateral forces possible Run-in time (approximately PT-1 behavior) - which is mainly a function of the rolling speed Phase delay of the simulator is theoretically depending z t on the current motion speed ψ t t 1 t 2 φ t θ t t 3 The tires rolling speed should always be on a same x t y t level F y 21.09.16 – Karlsruhe IPG Apply & Innovate ‘16 – Thomas Tüschen 9
3. auto.mobile-driving simulator Motion Concept ! End of driving maneuver – return into initial position Representation of driving Initial position maneuver ! Driving area Acting inertia force Drivers view (I-System) Rotation Velocity 21.09.16 – Karlsruhe IPG Apply & Innovate ‘16 – Thomas Tüschen 10
3. auto.mobile-driving simulator Motion Concept 21.09.16 – Karlsruhe IPG Apply & Innovate ‘16 – Thomas Tüschen 11
21.09.16 – Karlsruhe (Simulated vehicle) Car Maker CoG Vehicle acceleration/ angular velocity (6 DoF) Motion Filter Control System Overview 3. auto.mobile-driving simulator CoG IPG Apply & Innovate ‘16 – Thomas Tüschen Source: IPG Source: IPG Simulator acceleration/ angular velocity (7 DoF) Distribution Acceleration (COG) Force 4 forces (wheel-pair) based on wheel load Motion Control Control Force (wheel-pair) Force 2 drive torques (with related steering angles) 8 wheel torques; 4 steering angles auto.mobile-driving 4 Wheel forces x,y ; simulator slip angle (long., lat.) 4 Wheel loads Position 12
4. Prediction Phase delay CGI Car Maker & Driver Cockpit Drive Sim (Computer generated Control Sys. imagery) Driver input Cockpit input t tac Position physical Cockpit output Motion input objects time t tot Δ t kin Tactile stimuli Δ t vis Δ t kin Kinesthetic stimuli Visual/auditory stimuli 21.09.16 – Karlsruhe IPG Apply & Innovate ‘16 – Thomas Tüschen 13
4. Prediction Phase delay: CarMaker 2 Cars Host PC Master Driver Driver CarMaker for Simulink (Master) (Slave) UDP Communication Vehicle Vehicle (Master) (Slave) Coordinates for Ego Vehicle from Master Vehicle from Slave Coordinates for Ego UDP Environment Environment Communication (Master) (Slave) Slave CarMaker for Simulink Master tire model with Slave tire model relaxation length without relaxation F y F y length Δ t t t Source: IPG 21.09.16 – Karlsruhe IPG Apply & Innovate ‘16 – Thomas Tüschen 14
4. Prediction Phase delay: CarMaker 2 Cars Problems Results • UDP communication: • Time delay of about 10ms in yaw rate and lateral step size leads to a noisy signal acceleration • No synchronization of vehicle position: Results can be used as an indicator for predictive vehicle position of master and slave are drifting control system away 21.09.16 – Karlsruhe IPG Apply & Innovate ‘16 – Thomas Tüschen 15
4. Prediction Motion space Predictive model in vehicle simulation and Motion Filter Source: IPG Driving area Predictive Motion Filter [variable scaling 0,6; wash out acceleration max. 4 m/s², 0,85 m/s³] ! 21.09.16 – Karlsruhe IPG Apply & Innovate ‘16 – Thomas Tüschen 16
Thank you for your attention! Dipl.-Ing. Thomas Tüschen Tel.: +49 (0) 351/463-32831 Email: thomas.tueschen@tu-dresden.de URL: www.tu-dresden.de/kft Technische Universitaet Dresden Institute for Automotive Technologies Dresden Chair of Automotive Engineering George-Baehr-Strasse 1c 01062 Dresden Germany 21.09.16 – Karlsruhe IPG Apply & Innovate ‘16 – Thomas Tüschen 17
Recommend
More recommend
