With the development of smart factories, deep learning, which automatically extracts features and diagnoses faults, has become an important approach for fault diagnosis. In this paper, a novel interturn short-circuit fault (ISCF) diagnosis approach using an attention based recurrent neural network is proposed. An encoder-decoder architecture using an attention mechanism diagnoses the ISCF by estimating a fault indicator that directly reflects the severity of the fault, using currents and rotational speed signals as inputs. The attention mechanism helps the decoding process in accurate diagnosis and solves the long-term dependency problem of the encoder-decoder structure. The proposed algorithm uses only three-phase current and rotational speed as the inputs to evaluate the severity of the ISCF and enable early stage diagnosis of ISCF. The diagnosis of ISCF is achieved in various operating points and fault conditions, and no additional sensors such as voltage and vibration sensors are required. Experimental results for various operating and fault conditions demonstrate that the proposed method effectively diagnoses ISCFs.

Hydrogen energy storage (HES) has attracted renewed interest as a means to enhance the flexibility of power balancing to achieve the goal of a low-carbon grid. This paper presents an innovative data-driven HES model that reflects the interactive operations of an electrolyzer, a fuel cell, and hydrogen tanks. A model predictive control strategy is then developed, in which HES units support the frequency regulation (FR) of a microgrid (MG). In the proposed strategy, an MG-level controller is designed to optimize power sharing, to allow the HES units to respond quickly to power supply-and-demand imbalances, while distributed generators compensate for any remaining imbalance. The MG-level controller cooperates with the HES-level controllers, which change the operating modes and override the FR supports based on the hydrogen levels. Small-signal analysis is conducted to evaluate the contribution and sensitivity of the FR supports. Comparative case studies are also carried out, wherein HES model accuracy is verified and a hardware-in-the-loop simulation is implemented. The results of the small-signal analysis and case studies confirm that the proposed strategy is effective for reducing frequency deviations under various MG conditions, characterized by the net load demand, line congestion, plug-and-play, model parameters, and communication time delays.

Distributed state estimation (DSE) and bad data (BD) processing have become increasingly important for the analysis and operation of large-scale power networks. This paper develops a new strategy for decentralized phasor-aided state estimation (DPHASE) to improve the performance of BD processing and, hence, the accuracy of DSE results. Given the recent availability of PMUs, the DPHASE strategy is developed using a parallel multi-stage hybrid approach. The practical challenges in the approach are identified and attributed to the difference between the sets of buses for SCADA- and PMU-based state estimates in sub-areas. Phasor-aided normalized residuals are locally unobtainable and the fusion of local state estimates is unattainable. These challenges are effectively resolved by extending local state vectors. Comparative case studies are performed using the proposed and conventional DSE strategies. The study results confirm that the proposed DPHASE strategy improves the accuracy and robustness of the DSE results under various test network and data measurement conditions.

Network reconfiguration (NR) has received attention due to its potential to improve grid resilience by realizing self-healing microgrids (MGs). This paper proposes a new strategy for the real-time frequency regulation of a reconfigurable MG, wherein the feedforward control of synchronous and inverter-interfaced distributed generators (DGs) is achieved in coordination with the operations of sectionalizing and tie switches (SWs). This enables DGs to compensate more quickly, and preemptively, for a forthcoming variation in load demand due to NR-aided restoration. An analytical dynamic model of a reconfigurable MG is developed to analyze the MG frequency response to NR and determine the desired dynamics of the feedforward controllers when integrated with feedback loops for inertial response emulation and primary and secondary frequency control. A small-signal analysis is conducted to analyze the contribution of the feedforward control to the MG frequency regulation. Simulation case studies of NR-aided load restoration are also performed. The results of the small-signal analysis and case studies confirm that the proposed strategy is effective for the improvement of the MG frequency regulation under various conditions of load demand, model parameter errors, and communication time delays.