Layout of a Driver‐in‐the‐Loop Simulator for Validation of Human Body Models

Abstract

In the field of crash simulation, advanced human body models (HBMs) with muscle representation become more and more important because of their ability to emulate the behavior of real human occupants more accurately than dummy models. To get experimental validation data for these models concerning muscle activation and time-variant stiffness properties, a lowcost Driver-in-the-Loop (DiL) simulator based on a Porsche exhibition driving simulator is implemented. Driver kinematics and muscle activation are tracked by computer stereo vision and surface electromyography (SEMG). For improved driver immersion, kinematic feedback is provided by a six DoF motion platform, a force-feedback wheel and a bassshaker at the back of the driving seat. The DiL-simulation is based on a Simulink model, generated by the software package PreScan, with included mathematical models of vehicles, sensors, drivers, etc. The vehicle states, available in the generated Simulink model, are used to activate the motion platform via UDP protocol. Because of the physical limitations of the platform movement, the so-called Motion Cueing Algorithms (MCA) are necessary to emulate real driving sensation by providing motion cues which minimize the required translational platform movements.

Dynamic human body models in vehicle safety: An overview

“…The passive properties of the AHBM have been validated using PMHS tests [116]. Many active properties of the model have been indirectly validated by comparing the response of the entire head and neck model to in‐vivo sled tests [9,14,59], Driver‐in‐the‐Loop (DiL) tests [20,53,56], and tests in real vehicles [44,93]. An overview of available tests is provided by the OSCCAR consortium [90].…”

Dynamic Human Body Models in Vehicle Safety: An Overview

“…Madymo's control strategy is not hard-coded but can be customized by using the Matlab Madymo coupling strategy. For example, Kempter et al [50] implemented a completely different control strategy in which every individual muscle of the neck was controlled by its own stretch-reflex controller. By using the Matlab-Madymo coupling strategy, it is possible to update the HNC as more about this complex biological system is learned.…”

“…The Simcenter Madymo AHBM shows a lower numerical complexity than the FE models from the previous example. The calculation factor on Linux with Madymo R7.7 on a 12 core Intel i7-8700 processor with 3.20 GHz and 64 GB RAM is 21 min 4 s of calculation time for 2 s problem time, see [50].…”

“…DiL test allows to put the occupant into new (virtually) interior concepts while exposed to a mixed-reality scenario and to study the interaction under (limited) external loading. Different test series, see [50,53,20] were performed to gain a better understanding of human behavior in such situations. To analyze the driver's behavior in experiments, the kinematics and muscle activation are monitored, see Fig.…”

“…For the usage of more advanced or more flexible control strategies with the AHBM, the provided Matlab-Madymo coupling feature can be used. For example, in [50], an approach is presented to control the AHBM based on muscle-level controller, i.e., to model the monosynaptic stretch reflex [21], or muscle-length controller.…”