High cost and large volume are always associated with the conventional unified power flow controller (UPFC) formed by two back-to-back connected voltage source converters and two transformers. Both additional series-connected converters and their adjacent transformers are necessitated when controlling the power flow among multiple transmission lines. Such implementation will further increase the cost and volume of the equipment owing to the port expansion. In this regard, this paper proposes a multiport transformer-less UPFC (MTUPFC) with easy terminal expandability. The decoupled active and reactive power flow are also controlled independently. This device presents several advantages, including low cost, light weight, fast dynamic response characteristics, and easy port expansion. The MTUPFC therefore has significant advantages in power grids with complex configuration and large-scale renewable energy integration. The operation principle and control strategies of the proposed MTUPFC are elaborated. The feasibility and effectiveness of the MTUPFC are verified by a 60-MVA simulation model and a 2.5-kVA scaled-down experimental platform, respectively.

Condition monitoring (CM) has been considered as a promising technique to improve the reliability of IGBT. Among various condition parameters, switching time is a good health status indicator to detect IGBT failures. However, on-line monitoring of the IGBT high-speed switching time is still difficult in practice due to the requirement of the extremely high sampling rate for signal acquisition. To overcome the technical difficulty, this paper provides an innovative compressed sensing (CS) method to achieve equivalent sampling performance for high-speed IGBT switching time monitoring with lower sampling rate. By utilizing the sparse characteristics of IGBT switching signal, the sampling rate in the CS method could be far less than the traditional Nyquist sampling rate. To clarify the method, the CM mechanism using IGBT switching time is firstly analyzed. Then, three key points in the CS method are studied, i.e., the selection of sparsifying basis, the design of measurement matrix and the implementation of reconstruction algorithm. Finally, experiments are carried out to investigate the performance of the CS method for on-line CM. Experimental results show that the IGBT turn-off time at different temperatures can be accurately monitored with a highly compressed sampling rate, which validates the feasibility and effectiveness of the proposed method.

To improve the stability of a wind-diesel hybrid microgrid, a frequency control strategy is designed by using the hybrid energy storage system and the adjustable diesel generator with load frequency control (LFC). The objective of frequency control is to quickly respond to the disturbed system to reduce system frequency deviation and restore stability. By evaluating the area control error, the disturbance state of the system can be divided into four different areas by a corresponding control strategy for precise adjustments. For the diesel generator, an adaptive sliding mode (SM) algorithm is used to design LFC that can participate in frequency modulation. The frequency coordination control strategy proposed in this paper can realize the partition adjustment according to different resources, and ensure frequency stability. The proposed control strategy is verified by RTDS simulations in multiple scenarios.

The modular multilevel converter (MMC) open-circuit fault diagnosis under full power condition has become a challenging problem. Especially when MMC operates under low power condition, there are significant fluctuations in capacitor voltage and the existing fault diagnosis strategies face difficulties in selecting appropriate thresholds. The paper proposes a multiple open-circuit fault diagnosis strategy for MMC using Feature Reconstruction-Recurrent Neural Network (FR-RNN). In this method, the dominant characteristics of the theoretical and actual capacitor current of the submodules (SMs) are presented as the preferred feature set. The auxiliary mathematical feature based on capacitor current was screened using Maximal Information Coefficient (MIC), then the selected auxiliary mathematical feature and preferred feature set were inputted into the RNN for FR to realize fault diagnosis based on optimal feature set. The proposed strategy can directly achieve accurate fault localization and fault type judgment of SM faults without additional detection steps and diagnostic delay. Additionally, the determined optimal feature set greatly simplifies the network computation, thus avoiding deviation caused by multi-dimensional and redundant features on the diagnosis outputs. Simulation and experimental studies confirm the effectiveness of the strategy.

Real-time and high-precision extraction of groove types and key features is an important factor to achieve high performance of weld quality in the process of automatic welding. Based on the complexity and diversity of weld groove types, a method for identifying weld groove types (TL-Alexnet-ELM) is presented, which combines deep transfer learning with extreme learning machine. Firstly, to avoid over-fitting of the model, the weld data set is expanded using data enhancement technology. Then, to improve the generalization ability of the model, transfer learning is used to fine-tune the structure of Alexnet (TL-Alexnet) to improve the feature extraction accuracy of the source weld image. Finally, the extracted image eigenvectors are input into the extreme learning machine classifier to get the classification results of the groove types. To validate the effectiveness of the algorithm, model self-comparison study, comparison study of different deep learning network models, and comparison study of different classifiers are carried out. The experimental results show that the recognition accuracy of this method is 99.9%.