A Comparative Study On Spectrogram And S-Transform For Batteries Parameters Estimation
This research presents the analysis of battery charging and discharging signals using spectrogram, and S-transform techniques. The analysed batteries are lead acid (LA), nickel-metal hydride (Ni-MH), and lithium-ion (Li-ion). From the equivalent circuit model (ECM) simulated using MATLAB, the consta...
Saved in:
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Penerbit UTM Press
2019
|
| Online Access: | http://eprints.utem.edu.my/id/eprint/24628/2/12801-38678-1-PB.PDF http://eprints.utem.edu.my/id/eprint/24628/ https://journals.utm.my/jurnalteknologi/article/view/12801/6506 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | This research presents the analysis of battery charging and discharging signals using spectrogram, and S-transform techniques. The analysed batteries are lead acid (LA), nickel-metal hydride (Ni-MH), and lithium-ion (Li-ion). From the equivalent circuit model (ECM) simulated using MATLAB, the constant charging and discharging signals are presented, jointly, in time-frequency representation (TFR). From the TFR, the battery signal characteristics are determined from the estimated parameters of instantaneous means square voltage (V RMS (t)), instantaneous direct current voltage (V DC (t)), and instantaneous alternating current voltage (V AC (t)). Hence, an equation for battery remaining capacity as a function of estimated parameter of V AC (t) using curve fitting tool is presented. In developing a real-time automated battery parameters estimation system, the best time-frequency distribution (TFD) is chosen in terms of accuracy of the battery parameters, computational complexity in signal processing, and memory size. The advantages in high accuracy for battery parameters estimation, and low in memory size requirement makes the S-transform technique is selected to be the best TFD. Then, field testing is conducted for different cases, and the results show that the average mean absolute percentage error (MAPE) calculated is around 4%. |
|---|