measuring and analyzing road traffic noise - cityhush · 2012-12-17 · measuring and analyzing...
TRANSCRIPT
Measuring and Analyzing Road Traffic
Noise
Stockholm, December 11, 2012
CityHush Training Workshop
André Fiebig, HEAD acoustics GmbH
Stockholm, December 11, 2012 Fiebig: Road Traffic Noise 2
Structure
Traffic flow analysis with microphone arrays
and beamforming (measurements)
Traffic flow analysis with Traffic Noise
Synthesizer (simulations)
Stockholm, December 11, 2012 Fiebig: Road Traffic Noise 3
Traffic Flow Analysis with
Microphone Arrays
Stockholm, December 11, 2012 Fiebig: Road Traffic Noise 4
Objective of Work
Statistical Analysis of
vehicle driving behavior
and resulting noise
emission
• Vehicles?
• Speed?
• Acceleration?
• Noise?
• Sources? (wheels,
engine)
=
Real traffic flow Measurement
system
Stockholm, December 11, 2012 Fiebig: Road Traffic Noise 5
System Components
Video system
Detection of the vehicles
with position, speed, acceleration
Audio system
Localisation and quantification
of noise sources and emission
Panorama Video
Camera
with advanced image
processing software
Acoustic Camera
Microphone array with
beamforming software
Stockholm, December 11, 2012 Fiebig: Road Traffic Noise 6
Panorama Camera Module
Mercator projection 270
planar projection 160
• 3 VGA ethernet cameras: -60° , 0° , 60°
• Automated rectification (cancellation of
image aberrations,e.g. lense position
errors, fish eye)
• Automated stitching of panorama view
• All image errors (lense position, fish
eye) are corrected
Stockholm, December 11, 2012 Fiebig: Road Traffic Noise 7
Flexible Array
• Flexible array system
• Modular grids (1.5m x 1.5m)
• Up to 192 microphones
• Panorama camera module with
automated microphone position
detection
Stockholm, December 11, 2012 Fiebig: Road Traffic Noise 8
Optical Image = Vehicles + Background Acoustic Image = Sources
Available Data
Stockholm, December 11, 2012 Fiebig: Road Traffic Noise 9
Which Sources Correspond to Which
Vehicle?
• Continuous vehicle detection
• Vehicle flow simulation to compensate
detection errors
Stockholm, December 11, 2012 Fiebig: Road Traffic Noise 10
Vehicle Detection
(based on image processing)
background background with foreground
Foreground extraction detected vehicle
Stockholm, December 11, 2012 Fiebig: Road Traffic Noise 11
Traffic flow analysis - Example
Original video capture
Stockholm, December 11, 2012 Fiebig: Road Traffic Noise 12
Traffic flow analysis - Example
Detected objects
Stockholm, December 11, 2012 Fiebig: Road Traffic Noise 13
Traffic Flow Analysis - Example
Source level
detection by virtual
microphones
(related to 1m
distance)
Stockholm, December 11, 2012 Fiebig: Road Traffic Noise 14
Acoustical Fingerprint
of Citroen Electric Vehicle
Stockholm, December 11, 2012 Fiebig: Road Traffic Noise 15
Array Measurement of Electric Vehicle
Beamforming result
(Citroen electric vehicle)
Stockholm, December 11, 2012 Fiebig: Road Traffic Noise 16
Level vs. Constant Speed and
Frequency
Front tires Rear tires
low noise tires,
in
dB(A)
standard tires
Smooth
road surface
Stockholm, December 11, 2012 Fiebig: Road Traffic Noise 17
Level vs. Constant Speed and
Frequency Rough
road surface
standard tires
Front tires Rear tires
in
dB(A)
low noise tires
Stockholm, December 11, 2012 Fiebig: Road Traffic Noise 18
Conclusions I
• Measurements possible based on system with panorama camera
and flexible large scale array
• Optical detection of vehicles and calculation of physical parameters
• Acoustical detection of sources and generation of time signals
• Synthesis of large number of heterogenuous pass-by events
(acoustical fingerprint of single vehicles, specific road conditions )
• Data evaluation with standard and psychoacoustic analyses
Stockholm, December 11, 2012 Fiebig: Road Traffic Noise 19
Traffic Noise Synthesizer
Stockholm, December 11, 2012 Fiebig: Road Traffic Noise 20
Traffic Noise Synthesis of Vehicles
Motivation: Investigate effect of certain measures without
measurement uncertainties
Advantage:
Useable in subjective tests, calculation of any acoustic
parameter, audible forecasts possible (of virtual
modifications (e.g. warning signals)
Stockholm, December 11, 2012 Fiebig: Road Traffic Noise 21
Data flow of the traffic noise
synthesizer
Stockholm, December 11, 2012 Fiebig: Road Traffic Noise 22
Example of Synthesis:
Order analysis and synthesis
Spectrogram of a
near-field engine
measurement
Order analysis Order analysis
with detected orders
Re-synthesis of orders
f/Hz
100
200
500
1k
2k
5k
10k
20k
t/s5 6 7 8 9 11 12 13 15
L/dB[SPL]30 40 100O
/Ord
er
25
50
75
100
125
150
n/rpm3250 5000
L/dB[SPL]30 40 80 100
O/O
rder
25
50
75
100
125
150
n/rpm3250 5000
L/dB[SPL]30 40 100
f/Hz
100
200
500
1k
2k
5k
10k
20k
t/s1 2 3 4 6 7 8 9
L/dB[SPL]30 40 100
Stockholm, December 11, 2012 Fiebig: Road Traffic Noise 23
Example of Synthesis:
Sound synthesis of an electric vehicle
Order synthesis Noise synthesis Synthesis (superposition) Near-field measurement
f/Hz
20
50
100
200
500
1k
2k
5k
20k
t/s1 2 3 4 6 7 8 9
L/dB(A)[SPL]30 40 100
f/H
z
20
50
100
200
500
1k
2k
5k
10k
20k
t/s1 2 3 6 7 8 9
L/dB(A)[SPL]40 100
f/Hz
20
50
100
200
500
1k
2k
5k
20k
t/s1 2 3 4 6 7 8 9
L/dB(A)[SPL]30 40 100
f/Hz
20
50
100
200
500
1k
2k
5k
20k
t/s6 7 8 9 11 12 13 14
L/dB(A)[SPL]30 40 100
Stockholm, December 11, 2012 Fiebig: Road Traffic Noise 24
Sound example: Mitsubishi iMIEV
Single pass-by sound events during full-load
acceleration
iMIEV E-motor
iMIEV E-motor + tires
iMIEV E-motor + tires + converter
iMIEV E-motor + tires + converter + background noise
Stockholm, December 11, 2012 Fiebig: Road Traffic Noise 25
Traffic Noise Simulation
Traffic noise of a straight road (30 km/h)
with scooters and vehicles
Traffic noise scenario
with conventional drives Traffic noise scenario
with electric drives
Stockholm, December 11, 2012 Fiebig: Road Traffic Noise 26
Benefit of Traffic Noise Synthesizer
Identification of most efficient noise mitigation measures and
actions, which can be virtually experienced
Efficiency of measures can be assessed not only on the basis
of a dB(A)-reduction, but also on the reduction of further
relevant parameters
Decisions for or against specific noise mitigation measures are
more grounded
26
Stockholm, December 11, 2012 Fiebig: Road Traffic Noise 27
Andre Fiebig
www.head-acoustics.de
Thank you for your attention!