This paper proposes a novel intelligent fault diagnosis method to automatically identify different health conditions of wind turbine gearbox. Different from traditional approaches where feature extraction and classification are separately designed and performed, this study aims to automatically learns effective fault features directly from raw vibration signals and meanwhile classifies the type of faults in a single learning framework, thus leading to an end-to-end learning based fault diagnosis system for wind turbine gearbox without additional signal processing techniques and diagnostic expertise. Considering multi-scale characteristics inherent in vibration signals of a gearbox, in this paper, a new multiscale convolutional neural network (MSCNN) framework is proposed to perform multi-scale feature extraction and classification simultaneously. A key advantage of the proposed MSCNN is the ability to learn complementary and rich fault pattern features at different time scales by incorporating multiscale learning into the CNN architecture, which greatly improves feature learning to enable better diagnosis performance. The proposed MSCNN framework is evaluated through experiments on a wind turbine test rig. The results demonstrate that the proposed MSCNN framework outperforms state-of-the-art methods and exhibits superior robustness against noise.

Data-driven approaches have gained increasing interests in the fault detection of wind turbines (WTs) due to the difficulty in system modeling and the availability of sensor data. However, the nonlinearity of WTs, uncertainty of disturbances and measurement noise, and temporal dependence in time-series data still pose grand challenges to effective fault detection. To this end, this paper proposes a new fault detector based on a recently developed unsupervised learning method, denoising autoencoder (DAE), which offers the learning of robust nonlinear representations from data against noise and input fluctuation. A DAE is used to build a robust multivariate reconstruction model on raw time-series data from multiple sensors, and then, the reconstruction error of the DAE trained with normal data is analyzed for fault detection. In addition, we apply the sliding-window technique to consider temporal information inherent in time-series data by including the current and past information within a small time window. A key advantage of the proposed approach is the ability to capture the nonlinear correlations among multiple sensor variables and the temporal dependence of each sensor variable simultaneously, which significantly enhanced the fault detection performance. Simulated data from a generic WT benchmark and field supervisory control and data acquisition data from a real wind farm are used to evaluate the proposed approach. The results of two case studies demonstrate the effectiveness and advantages of our proposed approach.

In power systems, although the inertia energy in power sources can partly cover power unbalances caused by load disturbance or renewable energy fluctuation, it is still hard to maintain the frequency deviation within acceptable ranges. However, with the vehicle-to-grid (V2G) technique, electric vehicles (EVs) can act as mobile energy storage units, which could be a solution for load frequency control (LFC) in an isolated grid. In this paper, a LFC model of an isolated micro-grid with EVs, distributed generations and their constraints is developed. In addition, a controller based on multivariable generalized predictive control (MGPC) theory is proposed for LFC in the isolated micro-grid, where EVs and diesel generator (DG) are coordinated to achieve a satisfied performance on load frequency. A benchmark isolated micro-grid with EVs, DG, and wind farm is modeled in the Matlab/Simulink environment to demonstrate the effectiveness of the proposed method. Simulation results demonstrate that with MGPC, the energy stored in EVs can be managed intelligently according to LFC requirement. This improves the system frequency stability with complex operation situations including the random renewable energy resource and the continuous load disturbances.

The penetration of distributed energy resources (DERs) in smart grids significantly increases the number of field devices owned and controlled by consumers, aggregators, third parties, and utilities. As the interface between DER and power grids, DER inverters are becoming smarter with various grid-support functions and communication capabilities. Meanwhile, the cybersecurity risks of smart inverters are also on the rise due to the extensive utilization of information and communication technologies. The potential negative impacts of cyberattacks on smart inverters have attracted significant attention from scholars and organizations. To advance the research on smart inverter cybersecurity and provide insights into its technical achievements, barriers, and future directions, this article will give a comprehensive review of critical attacks and defense strategies for smart inverters and inverter-based systems like microgrids. We start this survey with an overview of the smart inverter introduction, including device- and grid-level architectures, grid-support functions, and communication protocols. We then review various cyberattacks and defense strategies in different categories and scenarios tailed with discussions including their feasibility and remaining gaps. Finally, we discuss the opportunities and challenges of emerging technologies that can secure smart inverters. We hope this survey can inspire efforts to close research gaps and develop more mature cybersecurity solutions for smart inverters in the smart grid.

The transition to electric vehicles (EVs) has prodigious plausibility in reducing green house gas (GHG). But EVs acceptance is, however, hindered by several challenges; among them is their avidity for quicker charging at lower price. This article considers a photovoltaic (PV)-powered station equipped with an energy storage system (ESS), which is assumed to be capable of assigning variable charging rates to different EVs to fulfill their demands inside their declared deadlines at minimum price. To ensure fairness, a charging rate-dependent pricing mechanism is proposed to assure a higher price for enjoying a higher charging rate. The PV generation profile and future load request are forecasted at each time slot, to handle the respective uncertainty. An integer linear programming (ILP)-based centralized system is first proposed to minimize the charging price per EV. Due to the larger computational time, we subsequently present two game theoretic algorithms, i.e., game 1 and game 2. In game 1, players are oblivious of upcoming charging requests, whereas in game 2, players consider the future anticipated load to select their charging strategies. The games are shown to converge to a Nash equilibrium. The average unit price of game 2 is shown to be the same as the one of the optimal solution and takes considerably less computation time than the centralized method.

Integrated energy systems (IES) with cooling, heat, electricity, and natural gas have drawn significant interest recently as we embrace more sustainable energy a midst climate change. However, the uncertain outputs of distributed generators (DGs) make it challenging for IES planning while maintaining low-cost installation and operation under carbon emission constraints. To tackle the challenge, this work proposes an optimal planning model for IES considering both DG output uncertainties and carbon emission punishments. To reduce the conservatism of the widely-adopted interval and affine algorithms, an affine model based on the matrix form is first proposed to model the uncertain DG outputs. A tiered dynamic charging cost model is further developed to introduce and minimize carbon emissions with a punishment mechanism at the planning stage. Based on these two sub-models, an optimal IES planning model is proposed to simultaneously minimize the overall costs of investment, operation, and carbon emissions. To solve the multi-dimensional nonlinear model, an improved quantum particle swarm optimization (IQPSO) algorithm is introduced with enhanced global optimization ability. Simulation results on the IEEE 33-bus and 69-bus IES network have demonstrated that the proposed method can effectively reduce the impacts of DG uncertainty and carbon emissions at the planning stage of IES with a better long-term economy.