The significant advance of power electronics in today’s market is calling for high-performance power conversion systems and MEMS devices that can operate reliably in harsh environments, such as high working temperature. Silicon-carbide (SiC) power electronic devices are featured by the high junction temperature, low power losses, and excellent thermal stability, and thus are attractive to converters and MEMS devices applied in a high-temperature environment. This paper conducts an overview of high-temperature power electronics, with a focus on high-temperature converters and MEMS devices. The critical components, namely SiC power devices and modules, gate drives, and passive components, are introduced and comparatively analyzed regarding composition material, physical structure, and packaging technology. Then, the research and development directions of SiC-based high-temperature converters in the fields of motor drives, rectifier units, DC–DC converters are discussed, as well as MEMS devices. Finally, the existing technical challenges facing high-temperature power electronics are identified, including gate drives, current measurement, parameters matching between each component, and packaging technology.

No abstract

No abstract

To improve the robustness and stability of the photovoltaic grid-connected inverter system, a nonlinear backstepping-based H ∞ controller is proposed. A generic dynamical model of grid-connected inverters is built with the consideration of uncertain parameters and external disturbances that cannot be accurately measured. According to this, the backstepping H ∞ controller is designed by combining techniques of adaptive backstepping control and L 2 -gain robust control. The Lyapunov function is used to design the backstepping controller, and the dissipative inequality is recursively designed. The storage functions of the DC capacitor voltage and grid current are constructed, respectively, and the nonlinear H ∞ controller and the parameter update law are obtained. Experimental results show that the proposed controller has the advantage of strong robustness to parameter variations and external disturbances. The proposed controller can also accurately track the references to meet the requirements of high-performance control of grid-connected inverters.INDEX TERMS Robustness, grid-connected inverter, H ∞ controller, adaptive backstepping control, L 2 -gain robust control.

Summary To develop a single‐stage power conversion topology in which energy storage devices can be directly coupled, a fluctuating voltage topology is applied, leading to lower cost and more compactness with the absence of DC/DC converters. This paper investigates such a topology for automotive applications where fuel cells are directly connected to the DC bus of the inverter, resulting in fluctuating voltage across the DC bus. Further, a supercapacitor pack is also introduced to maintain the power capacity and voltage stability. The hybridization principle and practical application of such a topology are then discussed in the time domain and frequency domain. Furthermore, the transient power requirement is decomposed to design the size of fuel cells and supercapacitors. Simulation results from the modeling of the fuel cell‐supercapacitor powertrain demonstrate the feasibility and effectiveness of this topology. The supercapacitors can serve as a low‐pass filter for the fuel cells. In conclusion, the peak power requirement can be successfully achieved because of the lowered system impedance, and the fuel cells only need to supply the average power.

Photovoltaic (PV) output is susceptible to meteorological factors, resulting in intermittency and randomness of power generation. Accurate prediction of PV power output can not only reduce the impact of PV power generation on the grid but also provide a reference for grid dispatching. Therefore, this paper proposes an LSTM-attention-embedding model based on Bayesian optimization to predict the day-ahead PV power output. The statistical features at multiple time scales, combined features, time features and wind speed categorical features are explored for PV related meteorological factors. A deep learning model is constructed based on an LSTM block and an embedding block with the connection of a merge layer. The LSTM block is used to memorize and attend the historical information, and the embedding block is used to encode the categorical features. Then, an output block is used to output the prediction results, and a residual connection is also included in the model to mitigate the gradient transfer. Bayesian optimization is used to select the optimal combined features. The effectiveness of the proposed model is verified on two actual PV power plants in one area of China. The comparative experimental results show that the performance of the proposed model has been significantly improved compared to LSTM neural networks, BPNN, SVR model and persistence model. INDEX TERMS LSTM-attention-embedding model, features extraction, deep learning, Bayesian optimization, residual connection.

No abstract

No abstract