Transcript
Modellering och p˚ averkan av rippelstr¨ommar i li-jonbatterier i tunga hybridfordon Oskar Wallmark1 , Rudi Soares1 , Alexander Bessman2 , G¨ oran 2 2,3 Lindbergh , and Pontus Svens 1 Institutionen 2 Avdelningen
f¨ or elkraftteknik, KTH
f¨ or till¨ ampad elektrokemi, KTH
3 Scania
AB, S¨ odert¨ alje
2017-05-18
Automotive electric drive
Figure 1: Electric drive (battery, three-phase converter and electric machine).
◮
Dc-link filter: Ibatt (ω) =
◮
1−
Iconv (ω) . 2 ω Lser C + jωCRser
Inductance Lser represents inductance due to cell arrangement and cable inductance.
Dc-link filter 1
Gain [-]
0.8 0.6 0.4 0.2 0 0
2000
4000
6000
8000
10000
f [Hz] Figure 2: C = 1 mF, Rser = 416 mΩ, Lser = 1 µH (blue), Lser = 10 µH (red).
Battery current harmonic content in a hybrid bus
◮
SL line 755 in S¨odert¨ alje.
◮
Tests carried out at test track at Scania (wide range of driving scenarios).
Battery current harmonic content in a hybrid bus
◮
Parallel drivetrain configuration.
◮
Five-cylinder ICE.
Battery current harmonic content in a hybrid bus
◮
Battery current and voltage sampled at 200 kHz.
◮
Simultaneous logging of CAN-bus data (50–100 Hz sampling frequency).
[km/h]
Battery current harmonic content in a hybrid bus 20 10
ibatt [A]
0 0
10
20
30
40
0
10
20
30
40
200 0 -200
t [s]
◮
Repeated acceleration and deceleration from (almost) zero speed.
◮
Both generative and regenerative action.
Battery current harmonic content in a hybrid bus 20
ibatt [A]
15 10 5 0 0
2000
4000
6000
8000
10000
f [Hz] ◮
Maximum value of magnitude of 0.2 s FFTs.
◮
Large magnitudes around ≈ 100 Hz (≈ 10% of maximum dc current), 2 kHz, 4 kHz, and 8 kHz.
◮
Small harmonics multiples of 100 Hz.
Battery current harmonic content in a hybrid bus
◮
A spectogram reveals the harmonic’s variation with time.
[A]
Battery current harmonic content in a hybrid bus 20 10 0 0
10
20
30
40
0
10
20
30
40
f [Hz]
200 100 0
t [s] ◮
Largest (in magnitude) harmonic as function of time.
◮
Blue line: measured data, red line: reproduced/fitted using CAN data only.
◮
Reproducing data using CAN data only ⇒ possibility to predict behaviour in standardized drive cycles.
An experimental setup for subjecting cells to ripple currents
An experimental setup for subjecting cells to ripple currents
◮
Pressure plates to avoid swelling.
◮
Heat sinks mounted to minimize temperature increase due to added ripple current.
An experimental setup for subjecting cells to ripple currents
◮
At most 16 cells can be cycled simultaneously.
◮
Climate chamber to increase temperature (40 ◦ C) (and potentially also humidity).
◮
Charge cycling with added sinusoidal current ripple.
◮
Automatic cell characterization.
Parallel activities/ongoing work
◮
Analysis of measurement data from second bus measurement (different vehicle with identical driveline).
◮
To what extent do the result from the two bus measurements differ?
Parallel activities/ongoing work
◮
◮
Temperature chamber allowing −70 to 100 ◦ C (room for single cell). Aim: Closed-loop thermal control of single cell by added (controlled) ac current.
Parallel activities/ongoing work
Figure 3: Measured impedance for two cells at −30 C◦ to 30 C◦ (SOC≈ 37%).
Summary ◮
This ongoing project focuses on the impact of ripple currents on battery cells. ◮
◮ ◮
Very preliminary data suggests that ... Low-frequency harmonics with significant magnitude (up to 10% of maximum dc-current level) was measured during in-vehicle tests. ◮
◮
◮
An experimental setup for subjecting multiple cells to ripple currents while charge cycling has been set up.
From the analysis of the first bus measurement, the magnitude and frequency could be accurately reproduced using data only from the CAN bus. Data from a second bus measurement yet to be analyzed.
Continued experiments planned until autumn 2018.
