The large-scale integration of renewable energy sources (RESs) and the increasing number of energy-storage (ES) systems connected to the grid will create a big challenge for power flow management at the distribution level. In this new scenario, a solid-state transformer (SST) will be a highly efficient key element for providing highly efficient and controllable power flow in distribution grids. SSTs not only mimic the operation of conventional transformers, which scale the voltage level between primary and secondary terminals, but they also control bidirectional power flow, stabilize voltage, facilitate direct integration of ES, and mitigate harmonics and transients, which can cause the power quality of the grid voltage to deteriorate. This article surveys several proposed structures for SSTs in the literature based on different ways of interconnecting power converter topologies.

Over the last decade, dual active bridge (DAB) converters have become critical components in high-frequency power conversion systems. Recently, intensive efforts have been directed at optimizing DAB converter design and control. In particular, several strategies have been proposed to improve the performance of DAB control systems. For example, fractional-order (FO) control methods have proven potential in several applications since they offer improved controllability, flexibility, and robustness. However, the FO controller design process is critical for industrializing their use. Conventional FO control design methods use frequency domain-based design schemes, which result in complex and impractical designs. In addition, several nonlinear equations need to be solved to determine the optimum parameters. Currently, metaheuristic algorithms are used to design FO controllers due to their effectiveness in improving system performance and their ability to simultaneously tune possible design parameters. Moreover, metaheuristic algorithms do not require precise and detailed knowledge of the controlled system model. In this paper, a hybrid algorithm based on the chaotic artificial ecosystem-based optimization (AEO) and manta-ray foraging optimization (MRFO) algorithms is proposed with the aim of combining the best features of each. Unlike the conventional MRFO method, the newly proposed hybrid AEO-CMRFO algorithm enables the use of chaotic maps and weighting factors. Moreover, the AEO and CMRFO hybridization process enables better convergence performance and the avoidance of local optima. Therefore, superior FO controller performance was achieved compared to traditional control design methods and other studied metaheuristic algorithms. An exhaustive study is provided, and the proposed control method was compared with traditional control methods to verify its advantages and superiority.

The smart home contributions in energy management systems can help the microgrid operator overcome technical problems and ensure economically viable operation by flattening the load profile. The purpose of this paper is to propose a smart home energy management system (SHEMS) that enables smart homes to monitor, store, and manage energy efficiently. SHEMS relies heavily on energy storage systems (ESSs) and electric vehicles (EVs), which enable smart homes to be more flexible and enhance the reliability and efficiency of renewable energy sources. It is vital to study the optimal operation of batteries in SHEMS; hence, a multi-objective optimization approach for SHEMS and demand response programs is proposed to simultaneously reduce the daily bills, the peak-to-average ratio, and the number of battery discharging cycles of ESSs and EVs. An inverter-based air conditioner, photovoltaic system, ESS, and EV, shiftable and non-shiftable equipment are considered in the suggested smart home. In addition, the amount of energy purchased and sold throughout the day is taken into account in the suggested mathematical formulation based on the real-time market pricing. The suggested multi-objective problem is solved by an improved gray wolf optimizer, and various weather conditions, including rainy, sunny, and cloudy days, are also analyzed. Additionally, simulations indicate that the proposed method achieves optimal results, with three objectives shown on the Pareto front of the optimal solutions.

Finite control set model predictive control (FCSMPC) offers several benefits for power electronic applications, including fast response, precise tracking, and high performance. However, reliance on an accurate model and its parameters has hindered the advancement of FCSMPC methods in power electronics inverter systems due to the uncertain model and parameter variations of the load and filter side. Recently, model-free FCSMPC (MF-FCSMPC) methods have been presented in the literature to determine the model parameters of the employed auto-regressive with exogenous input (ARX) model of the output resistance-inductor (R-L) load side. The majority of existing MF-FCSMPC methods employ the recursive least squares (RLSs) algorithm to update parameters in real-time. However, the conventional RLS method critically suffers from numerical instability and high computational cost, which limits its application in practical power inverter systems. Additionally, the performance of the RLS-based MF-FCSMPC methods relies on the initial model of the system. To overcome this issue, this article presents an improved MF-FCSMPC method based on the affine projection (AP) algorithm, which guarantees numerical stability and low computational cost. The proposed AP-based MF-FCSMPC is less sensitive to initial model parameters. The experimental results of the proposed AP-based MF-FCSMPC method demonstrate its ability to perform fast control using FCSMPC without prior knowledge of the physical plant, compared to classical model-based FCSMPC.

The Differential Boost Inverter (DBI) is a single-stage step-up topology composed of two bidirectional boost converters whose outputs are connected in differential mode to generate an ac output voltage. The DBI poses significant modeling challenges, since all variables operate with both dc and ac components, requiring a highly complex dynamical model to accurately capture its behavior. Also, the operating range of this converter is restricted both in the modulation indexes, which is limited in the interval between 0 and 1, and in the capacitor voltage which must always be above the input voltage. In addition, a large harmonic content in the converter variables when working at certain operating points can increase the Total Harmonic Distortion (THD), degrading the performance. This work presents a model that considers both the dc and fundamental ac components of the converter variables. Based on this model, the allowable operating ranges for the modulation indexes and capacitor voltages are defined. Furthermore, the model is extended to determine the harmonic content of the capacitor voltage enabling the identification of the operating ranges in which a low THD is achieved. The proposed analysis introduces a flexible dc component in the modulation, without assuming a fixed dc bias, which broadens the valid operating region and enhances the potential for THD minimization. Finally, a DBI prototype has been built to validate the model and analysis, with experimental results confirming the mathematical relationships at representative operating points.