Content uploaded by Andor Fürjes
Author content
All content in this area was uploaded by Andor Fürjes on Aug 02, 2018
Content may be subject to copyright.
Adaptive Automotive Acoustic
Emission Control
Introduction
This presentation was written for a kick off meeting in February
of 2016. The aim was to show some ideas of upcoming challenges
in acoustic emission control of vehicles with silent powertrains
and possible uses of inverse acoustic modeling methods.
I share this presentation with the hope that some of my thoughts
are not obsolete yet. Yeap, it already has passed 30 months since
then…
Csákvár, 2018
Andor T. Fürjes
1
Problem: contradictory aims in acoustic design of automotive vehicles.
acoustic emmission from
transportation shall be
minimized
safety requires acoustic signaling to
both inside and outside vehicles
cabin noise and vibration
shall be minimized for
better comfort and less
fatigue acoustic feedback from ride
increases the enjoyment
Adaptive Automotive Acoustic Emission Control
Andor T. Fürjes, 2016
2
Importance: less combustion engine, less ‘native’ acoustic emission.
Trend forecast: modular electric car
•parts of the chassis are less dependent on each other (e.g. drive by wire)
•configuration of main parts is user selectable (require standardized
mechanical and electric interfacing of parts)
•cabin interior and chassis shell/envelope more individualized (3D printing).
Consequences for a modular electric car:
•more affordable NVH and virtual prototyping required (more variations);
•acoustic emission must be controlled (safety vs. less noise);
•not only damping and attenuation but active acoustic emission is also
necessary;
•individualization and acoustic branding will be much more important.
Adaptive Automotive Acoustic Emission Control
Andor T. Fürjes, 2016
3
Solution:
•loudspeakers are used to emit sound to the environment
•loudspeakers are used to emit sound to the cabin
•acoustic actuators (shakers) are used for controlled vibration in the cabin
•acoustic sensors are required
Such a system must be adaptive:
•ride situation (velocity, acceleration, parking)
•environment (city, highway, etc.)
•traffic dependent (nearby other vehicles)
Adaptive Automotive Acoustic Emission Control
Andor T. Fürjes, 2016
4
DSP
(signal transfer
matrix)
loudspeakers
vibroacoustic actuators
drive data
(velocity, acceleration, etc.)
navigation
(GPS)
environmental scanners
(proximity sensor, etc.)
wireless inter-vehicle
communication
acoustic
synthesizer
ultrasound emitter
personal settings
multimedia
Adaptive Automotive Acoustic Emission Control
Andor T. Fürjes, 2016
5
Example: road traffic situations, artificial sound emission
•town: emission to only pedestrians (beam steering)
•highway: ultra sound emission only (beam steering)
•cross roads: raised sonic signalling if zebra crossing detected
•pedestrial warning: differentiation of sonic signature to signal direction of
movement (approaching, receding), speed, acceleration, braking, etc.
•parking (reverse drive, etc.)
Task:
•sound design: emitted sound shall reflect direction, weight, speed and
acceleration as well as individual/branded characteristics
•excess interference with other vehicles shall be avoided
•legislation required: acoustic signaling vs. noise regulations
•adaptive: direction and level of emitted sound
Adaptive Automotive Acoustic Emission Control
Andor T. Fürjes, 2016
6
Example: road traffic situations, cabin sound
•driving: ‘engine’ sound emphasized at driver position only
•autonomuous drive: low level signalling of what the car will do
•simultaneous multimedia: talk (masking or voice lift), phone call, watching
videos, etc. emphasized at the required position
•emergency situation: sound image reflects direction of upcoming emergency (e.g.
slow approaching obstacle, fast approaching ambulance)
Task:
•sound design: cabin sound (‚engine sound’)
•no active cancellation but rather creating ‚sweet spots’ (see wave field synthesis)
•adaptive: lower levels during talk or phone calls
•legislation, common platforms
Adaptive Automotive Acoustic Emission Control
Andor T. Fürjes, 2016
7
Problem:
•how is the DSP transfer matrix set up?
•optimization through series of measurements is not enough: set up must be
adaptive
•individualization (see modular design, 3D printed cabin, etc.)
Solution: automated optimization through inverse acoustic modelling in
order to understand
•the minimum number of actuators/loudspeakers required;
•to optimize location of actuators;
•to synthesize acoustic wave fields around the vehicle and in the cabin of the
vehicle by parametric modelling of optimization results (using interpolation to
smooth transition when changing between modelled settings)
Adaptive Automotive Acoustic Emission Control
Andor T. Fürjes, 2016
8
What is: inverse acoustic modelling?
The usual/common way of using models for optimization: ‘trial and error’ or
random gradient optimization (genetic optimization is a fancy variant)
MODEL
model settings
and parameters
model
results
let’s try again if results are not satisfactory:
new set of model settings and parameters
The ‘inverse’ modeling way of optimization: no iteration but real analysis
MODEL
model settings
and parameters
model
results
Adaptive Automotive Acoustic Emission Control
Andor T. Fürjes, 2016
9
Our acoustic environment
… is a resource, because we all share the same wave field
Adaptive Automotive Acoustic Emission Control
Andor T. Fürjes, 2016
… must be used and protected in a controlled and smart way.
10
- 11 -
Adaptive Automotive Acoustic Emission Control
Andor T. Fürjes, 2016