Reduced‐order modeling of lead‐acid battery using cluster analysis and orthogonal cluster analysis method

Abstract

Summary Real‐time fast simulation of lead‐acid battery (LAB) plays an important role in monitoring, control, optimization, and many other engineering fields. Hence, any improvement toward a reduction in computational time of LAB simulation while maintaining the accuracy of results is of practical interest. Reduced‐order modeling (ROM) is one of the promising tools, which is computationally cost‐effective along with producing accurate results. In this study, ROM is employed in a transient one‐dimensional simulation of LABs within discharge to investigate the variation of battery parameters, eg, cell voltage, acid concentration, and state of charge (SoC). Accordingly, three reduced‐order models are implemented, namely, proper orthogonal decomposition (POD), cluster analysis (CA), and orthogonal cluster analysis (OCA), wherein the latter one is a new hybrid model of POD and CA methods proposed in the present work. The results reveal that ROM of LAB reduces the simulation time significantly (speed‐up factor about 7‐12) and provide good consistency comparing with previous experimental and numerical studies (less than 1% relative error for cell voltage, acid concentration, and SoC). In addition, the results indicate that the new hybrid method inherits the advantages of both POD and CA methods, ie, enhances the speed of POD method by 8% to 24% and accuracy of CA method by 17% to 65%.

Lithium‐ionbatterySOC/SOHadaptive estimation via simplified single particle model

“…Recently, lithium‐ion battery modeling and simulation research became a hot topic in both academics and industry. The primary battery modeling approach is Equivalent Circuit Models (ECM), 5,20‐25 which can be classified as Internal Resistance (RINT) models, Thevinnen models, Partnership for a New Generation of Vehicle (PNGV) models, Resistance Capacitance (RC) models, and Dual Polarization (DP) models in terms of circuit topologies. The ECM model has the advantage of simplified electrical representation and is general to all types of battery.…”

Experimental investigation of the effect of C‐rate, electrode gaps, and electrode surface roughness on the performance characterization of lead‐acid batteries

“…It should be noted that so far many researchers have tried to numerical modeling of the thermo-electrochemical behavior of the battery to investigate the LAB performance under different conditions. [22][23][24][25][26][27][28] The first comprehensive model to simulate the behavior of a LAB in the discharge process was presented by Newmann and Tiedemann. 22 Subsequent efforts by Sunu et al, 23 Go et al, 24,25 Esfahanian et al, 26,27 and Shahbazi and Esfahanian 28 have led to the improvement of these numerical models.…”

“…[22][23][24][25][26][27][28] The first comprehensive model to simulate the behavior of a LAB in the discharge process was presented by Newmann and Tiedemann. 22 Subsequent efforts by Sunu et al, 23 Go et al, 24,25 Esfahanian et al, 26,27 and Shahbazi and Esfahanian 28 have led to the improvement of these numerical models. It seems that these numerical models still need improvement due to simplifications in numerical modelings, such as simplification of equations and determination of the thermo-physical properties affecting the thermo-electrochemical process of the FLAB.…”

“…Therefore, one of the indicators of performance improvement in FLAB is the increased capacity in the fast charging and discharging processes with the lowest TR. To this goal, many researchers have investigated various influential parameters such as the composition of alloy composites, the structure and geometry of FLAB, the electrode grid, ambient temperature, and pressure conditions and the additives to the aqueous electrolyte solution, electrolyte flow velocity, and so forth 5‐28 . Among these are Pavlov's extensive efforts to study the various combinations of active materials of negative electrodes in the past two decades, some of which are presented in the literature 5 .…”

“…Nguyen and White 21 investigated the effect of distance between positive and negative electrodes and the thickness of the separator on the performance and capacity of the LAB cells and showed that the behavior of the LAB capacity is a linear relationship with thickness of the separator. It should be noted that so far many researchers have tried to numerical modeling of the thermo‐electrochemical behavior of the battery to investigate the LAB performance under different conditions 22‐28 . The first comprehensive model to simulate the behavior of a LAB in the discharge process was presented by Newmann and Tiedemann 22 .…”