Electrical models of battery cells are used in simulations to represent batteries’ behavior in various fields of research and development involving battery cells and systems. Electrical equivalent circuit models, either linear or nonlinear, are commonly used for this purpose and are presented in this article. Various commercially available cylindrical, state-of-the-art lithium-ion battery cells, both protected and unprotected, are considered. Their impedance properties, according to four different equivalent circuit models, are measured using electrochemical impedance spectroscopies. Furthermore, the pricing, impedance, specific energy, and C-rate of the chosen battery cells are compared. For example, it is shown that the energy density of modern 18650 cells can vary from a typical value of 200 to about 260 Wh kg⁻¹, whereas the cell price can deviate by a factor of about 3 to 5. Therefore, as a result, this study presents a concise but comprehensive battery parameter library that should aid battery system designers or power electronic engineers in conducting battery simulations and in selecting appropriate battery cells based on application-specific requirements. In addition, the accuracies and computational efforts of the four equivalent circuit models are compared.

The use of renewable energy sources such as solar photovoltaic, wind, and fuel cells is becoming increasingly prevalent due to a combination of environmental concerns and technological advancements, as well as decreasing production costs. Power electronics DC-DC converters play a key role in various applications, including hybrid energy systems, hybrid vehicles, aerospace, satellite systems, and portable electronic devices. These converters are used to convert power from renewable sources to meet the demands of the load, improving the dynamic and steady-state performance of green generation systems. This study presents a comparison of the most commonly used non-isolated DC-DC converters for fuel cell applications. The important factors considered in the comparison include voltage gain ratio, voltage switch stress, voltage ripple, efficiency, cost, and ease of implementation. Based on the comparison results, the converters have been grouped according to voltage level applications, with low voltage applications being best served by converters such as DBC, DuBC, TLBC, 2-IBC, 1st M-IBC, PSOL, SEPIC, and 1st M-SEPIC owing to their lower cost, smaller size, and reduced switch stress. Medium voltage applications are best suited to converters such as TBC, 1st M-TLBC, 2nd M-TLBC, 4-IBC, 1st M-IBC, 2nd M-IBC, 1st M-PSOL, 2nd M-PSOL, 1st M-SEPIC, and 2nd M-SEPIC, which offer higher efficiency. Finally, high voltage applications are best served by converters such as TBC, 1st M-TBC, 2nd M-IBC, 3rd M-IBC, 3rd M-PSOL, 4th M-PSOL, 2nd M-SEPIC, 3rd M-SEPIC, and 4th M-SEPIC.

This paper investigates the potential of bidirectional charging using modular multilevel inverter-based reconfigurable battery systems via grid-parallel control. The system offers several advantages such as modularity, scalability, and fault-tolerance over conventional battery electric vehicle systems. It is designed for seamless integration with the grid, allowing bidirectional power flow and efficient energy storage. Within this study, the battery system is first simulated in Matlab/Simulink and later implemented into a hardware setup. Eventually, the simulation results and the measurements have been compared and evaluated. Thereby, startup sequences and constant current scenarios were investigated. It has been shown that the system is fully capable to charge and discharge the batteries in the grid-parallel connection, thus enabling bidirectional charging with close to full drive system power. The current total harmonic distortion complies with grid regulations and can potentially improve the grid quality. The proposed system offers significant potential for grid-integrated energy storage systems, addressing the challenges associated with renewable energy integration, grid stability, and energy management. In comparison to other publications on this topic, the proposed approach does not need additional dedicated power electronic hardware and has more degrees of freedom for current control.

This paper presents a novel capacitor voltage balancing control approach for cascaded multilevel inverters with an arbitrary number of series-connected H-Bridge modules (floating capacitor modules) with asymmetric voltages, tiered by a factor of two (binary asymmetric). Using a nearest-level reference waveform, the balancing approach uses a one-step-ahead approach to find the optimal switching-state combination among all redundant switching-state combinations to balance the capacitor voltages as quickly as possible. Moreover, using a Lyapunov function candidate and considering LaSalle’s invariance principle, it is shown that an offline calculated trajectory of optimal switching-state combinations for each discrete output voltage level can be used to operate (asymptotically stable) the inverter without measuring any of the capacitor voltages, achieving a novel sensorless control as well. To verify the stability of the one-step-ahead balancing approach and its sensorless variant, a demonstrator inverter with 33 levels is operated in grid-tied mode. For the chosen 33-level converter, the NPC main-stage and the individual H-bridge modules are operated with an individual switching frequency of about 1 kHz and 2 kHz, respectively. The sensorless approach slightly reduced the dynamic system response and, furthermore, the current THD for the chosen operating point was increased from $3.28$ to $4.58$ in comparison with that of using the capacitor voltage feedback.

The battery-based multilevel inverter has grown in popularity due to its ability to boost a system’s safety while increasing the effective battery life. Nevertheless, the system’s high degree of freedom, induced by a large number of switches, provides difficulties. In the past, central computation systems that needed extensive communication between the master and the slave module on each cell were presented as a solution for running such a system. However, because of the enormous number of slaves, the bus system created a bottleneck during operation. As an alternative to conventional multilevel inverter systems, which rely on a master–slave architecture for communication, decentralized controllers represent a feasible solution for communication capacity constraints. These controllers operate autonomously, depending on local measurements and decision-making. With this approach, it is possible to reduce the load on the bus system by approximately 90 percent and to enable a balanced state of charge throughout the system with an absolute maximum standard deviation of $1.1\times {10}^{-5}$. This strategy results in a more reliable and versatile multilevel inverter system, while the load on the bus system is reduced and more precise switching instructions are enabled.

Lithium-ion batteries are a key technology for the electrification of the transport sector and the corresponding move to renewable energy. It is vital to determine the condition of lithium-ion batteries at all times to optimize their operation. Because of the various loading conditions these batteries are subjected to and the complex structure of the electrochemical systems, it is not possible to directly measure their condition, including their state of charge. Instead, battery models are used to emulate their behavior. Data-driven models have become of increasing interest because they demonstrate high levels of accuracy with less development time; however, they are highly dependent on their database. To overcome this problem, in this paper, the use of a data augmentation method to improve the training of artificial neural networks is analyzed. A linear regression model, as well as a multilayer perceptron and a convolutional neural network, are trained with different amounts of artificial data to estimate the state of charge of a battery cell. All models are tested on real data to examine the applicability of the models in a real application. The lowest test error is obtained for the convolutional neural network, with a mean absolute error of 0.27%. The results highlight the potential of data-driven models and the potential to improve the training of these models using artificial data.