Explore the vast range of titles available.

Unbalanced active powers can affect power quality and system reliability due to high penetration and uneven allocation of single-phase photovoltaic (PV) rooftop systems and load demands in a three-phase four-wire microgrid. This paper proposes a distributed control strategy to alleviate the unbalanced active powers using distributed single-phase battery storage systems. In order to balance the unbalanced active powers at the point of common coupling (PCC) in a distributed manner, the agents (households’ single-phase battery storage systems) must have information on the active powers and phases. Inspired by supervised learning, a clustering approach was developed to use labels in order to match the three-phase active powers at the PCC with the agents’ phases. This enables the agent to select the correct active power data from the three-phase active powers. Then, a distributed power balancing control strategy is applied by all agents to compensate the unbalanced active powers. Each agent calculates the average grid power based on information received from its neighbours so that all agents can then cooperatively operate in either charging or discharging modes to achieve the compensation. As an advantage, the proposed distributed control strategy offers the battery owners flexibility to participate in the strategy. Case studies comparing performance of local, centralized, and the proposed distributed strategy on a modified IEEE-13-bus test system with real household PV powers and load demands are provided.

The penetration and integration of renewable energy sources into modern power systems has been increasing over recent years. This can lead to frequency excursion and low inertia due to renewable energy sources’ intermittency and absence of rotational synchronous machines. Battery energy storage systems can play a crucial role in providing the frequency compensation because of their high ramp rate and fast response. In this paper, a decentralized frequency control system composed of three parts is proposed. The first part provides adaptive frequency droop control with its droop coefficient a function of the real-time state of charge of battery. The second part provides a fully decentralized frequency restoration. In the third part, a virtual inertia emulation improves the microgrid resilience. The presented results demonstrate that the proposed control system improves the microgrid resilience and mitigates the frequency deviation when compared with conventional ω -P droop control and existing control systems. The proposed control system is verified on Real-Time Digital Simulator (RTDS), with accurate microgrid model, nonlinear battery models and detailed switching models of power electronic converters.

Recently, deep learning techniques have become popular and are widely employed in several research areas, such as optimization, pattern recognition, object identification, and forecasting, due to the advanced development of computer programming technologies. A significant number of renewable energy sources (RESs) as environmentally friendly sources, especially solar photovoltaic (PV) sources, have been integrated into modern power systems. However, the PV source is highly fluctuating and difficult to predict accurately for short-term PV output power generation, leading to ineffective system planning and affecting energy security. Compared to conventional predictive approaches, such as linear regression, predictive-based deep learning methods are promising in predicting short-term PV power generation with high accuracy. This paper investigates the performance of several well-known deep learning techniques to forecast short-term PV power generation in the real-site floating PV power plant of 1.5 MWp capacity at Suranaree University of Technology Hospital, Thailand. The considered deep learning techniques include single models (RNN, CNN, LSTM, GRU, BiLSTM, and BiGRU) and hybrid models (CNN-LSTM, CNN-BiLSTM, CNN-GRU, and CNN-BiGRU). Five-minute resolution data from the real floating PV power plant is used to train and test the deep learning models. Accuracy indices of MAE, MAPE, and RMSE are applied to quantify errors between actual and forecasted values obtained from the different deep learning techniques. The obtained results show that, with the same training dataset, the performance of the deep learning models differs when testing under different weather conditions and time horizons. The CNN-BiGRU model offers the best performance for one-day PV forecasting, while the BiLSTM model is the most preferable for one-week PV forecasting.

With high penetration of renewable energy sources (RESs), advanced microgrid distribution networks are considered to be promising for covering uncertainties from the generation side with demand response (DR). This paper analyzes the effectiveness of multi-objective optimization in the optimal resource scheduling with consumer fairness under renewable generation uncertainty. The concept of consumer fairness is considered to provide optimal conditions for power gaps and time gaps. At the same time, it is used to mitigate system peak conditions and prevent creating new peaks with the optimal solution. Multi-objective gray wolf optimization (MOGWO) is applied to solve the complexity of three objective functions. Moreover, the best compromise solution (BCS) approach is used to determine the best solution from the Pareto-optimal front. The simulation results show the effectiveness of renewable power uncertainty on the aggregate load profile and operation cost minimization. The results also provide the performance of the proposed optimal scheduling with a DR program in reducing the uncertainty effect of renewable generation and preventing new peaks due to over-demand response. The proposed DR is meant to adjust the peak-to-average ratio (PAR) and generation costs without compromising the end-user’s comfort.

Thailand’s policies are in accord with the global drive to electrify transportation vehicle fleets due to climate concerns. This dedication is evident through its adoption of the 30@30 initiative and the planned ban on new internal combustion (IC) engine vehicles by 2035, showcasing a strong commitment. The objective of this study was to utilize the Low Emission Analysis Platform (LEAP) software to model the transition possibilities for electric vehicle (EV). Emphasis was placed on the future of the light-duty vehicle (LDV) sector, encompassing the energy sources, electric power demands, and greenhouse gas (GHG) emissions. Two scenarios were evaluated: one involving rapid economic growth and the other characterized by a more-gradual expansion. The former projection foresees 382 vehicles per thousand people by 2040, while the latter estimate envisions 338 vehicles. In the scenario of high growth, the vehicle stock could surge by 70% (27-million), whereas in the case of low growth, it might experience a 47% rise (23.3-million) compared to the base year (15.8 million). The increased adoption of EVs will lead to a decrease in energy demand owing to improved fuel efficiency. Nonetheless, even in the most-extreme EV scenarios, the proportion of electricity in the energy mix will remain below one-third. While GHG emissions will decrease, there is potential for even greater emission control through the enforcement of stricter emission standards. Significant EV adoption could potentially stress power grids, and the demand for charging might give rise to related challenges. The deployment of public fast charging infrastructure could provide a solution by evenly distributing the load across the day. In the most-rapid EV penetration scenario, a public charging program could cap the demand at 9300 MW, contrasting with the 21,000 MW demand for home charging. Therefore, a recommended approach involves devising an optimal strategy that considers EV adoption, a tariff structure with incentives, and the preparedness of the infrastructure.

In standalone microgrids, the Battery Energy Storage System (BESS) is a popular energy storage technology. Because of renewable energy generation sources such as PV and Wind Turbine (WT), the output power of a microgrid varies greatly, which can reduce the BESS lifetime. Because the BESS has a limited lifespan and is the most expensive component in a microgrid, frequent replacement significantly increases a project’s operating costs. This paper proposes a capacity optimization method as well as a cost analysis that takes the BESS lifetime into account. The weighted Wh throughput method is used in this paper to estimate the BESS lifetime. Furthermore, the well-known Particle Swarm Optimization (PSO) algorithm is employed to maximize battery capacity while minimizing the total net present value. According to simulation results, the optimal adjusting factor of 1.761 yields the lowest total net present value of US$200,653. The optimal capacity of the BESS can significantly reduce the net present value of total operation costs throughout the project by extending its lifetime. When applied to larger power systems, the proposed strategy can further reduce total costs.

As the integration of renewable energy sources (RESs) and distributed generations (DGs) increases, the need for stable and reliable operation of microgrids (MGs) becomes crucial. However, the inherent low inertia of such systems poses intricate control challenges that necessitate innovative solutions. To tackle these issues, this paper presents the development of robust-adaptive controllers tailored specifically for grid-forming (GFM) converters. The proposed adaptive-robust controllers are designed to accommodate the diverse range of scenarios encountered in low-inertia MGs. The proposed approach applies both the robust control techniques and adaptive control strategies, thereby offering an effective means to ensure stable and seamless converter performance under varying operating conditions. The efficacy of the introduced adaptive-robust controllers for GFM converters is validated within a low-inertia MG, which is characterized by substantial penetration of converter-interfaced resources. The validation also encompasses diverse MG operational scenarios and conditions.

Smart energy management and control systems can improve the efficient use of electricity and maintain the balance between supply and demand. This paper proposes the modeling of a decentralized energy management system (EMS) to reduce system operation costs under renewable generation and load uncertainties. There are three stages of the proposed strategy. First, this paper applies an autoregressive moving average (ARMA) model for forecasting PV and wind generations as well as power demand. Second, an optimal generation scheduling process is designed to minimize system operating costs. The well-known algorithm of particle swarm optimization (PSO) is applied to provide optimal generation scheduling among PV and WT generation systems, fuel-based generation units, and the required power from the main grid. Third, a demand response (DR) program is introduced to shift flexible load in the microgrid system to achieve an active management system. Simulation results demonstrate the performance of the proposed method using forecast data for hourly PV and WT generations and a load profile. The simulation results show that the optimal generation scheduling can minimize the operating cost under the worst-case uncertainty. The load-shifting demand response reduced peak load by 4.3% and filled the valley load by 5% in the microgrid system. The proposed optimal scheduling system provides the minimum total operation cost with a load-shifting demand response framework.

Because of environmentally friendly reasons and advanced technological development, a significant number of renewable energy sources (RESs) have been integrated into existing power networks. The increase in penetration and the uneven allocation of the RESs and load demands can lead to power quality issues and system instability in the power networks. Moreover, high penetration of the RESs can also cause low inertia due to a lack of rotational machines, leading to frequency instability. Consequently, the resilience, stability, and power quality of the power networks become exacerbated.This thesis proposes and develops new strategies for energy storage (ES) systems distributed in power networks for compensating for unbalanced active powers and supply-demand mismatches and improving power quality while taking the constraints of the ES into consideration. The thesis is mainly divided into two parts.In the first part, unbalanced active powers and supply-demand mismatch, caused by uneven allocation and distribution of rooftop PV units and load demands, are compensated by employing the distributed ES systems using novel frameworks based on distributed control systems and deep reinforcement learning approaches.There have been limited studies using distributed battery ES systems to mitigate the unbalanced active powers in three-phase four-wire and grounded power networks. Distributed control strategies are proposed to compensate for the unbalanced conditions. To group households in the same phase into the same cluster, algorithms based on feature states and labelled phase data are applied. Within each cluster, distributed dynamic active power balancing strategies are developed to control phase active powers to be close to the reference average phase power. Thus, phase active powers become balanced.To alleviate the supply-demand mismatch caused by high PV generation, a distributed active power control system is developed. The strategy consists of supply-demand mismatch and battery SoC balancing. Control parameters are designed by considering Hurwitz matrices and Lyapunov theory. The distributed ES systems can minimise the total mismatch of power generation and consumption so that reverse power flowing back to the main is decreased. Thus, voltage rise and voltage fluctuation are reduced.Furthermore, as a model-free approach, new frameworks based on Markov decision processes and Markov games are developed to compensate for unbalanced active powers. The frameworks require only proper design of states, action and reward functions, training, and testing with real data of PV generations and load demands. Dynamic models and control parameter designs are no longer required. The developed frameworks are then solved using the DDPG and MADDPG algorithms.In the second part, the distributed ES systems are employed to improve frequency, inertia, voltage, and active power allocation in both islanded AC and DC microgrids by novel decentralized control strategies.In an islanded DC datacentre microgrid, a novel decentralized control of heterogeneous ES systems is proposed. High- and low-frequency components of datacentre loads are shared by ultracapacitors and batteries using virtual capacitive and virtual resistance droop controllers, respectively. A decentralized SoC balancing control is proposed to balance battery SoCs to a common value. The stability model ensures the ES devices operate within predefined limits.In an isolated AC microgrid, decentralized frequency control of distributed battery ES systems is proposed. The strategy includes adaptive frequency droop control based on current battery SoCs, virtual inertia control to improve frequency nadir and frequency restoration control to restore system frequency to its nominal value without being dependent on communication infrastructure. A small-signal model of the proposed strategy is developed for calculating control parameters.The proposed strategies in this thesis are verified using MATLAB/Simulink with Reinforcement Learning and Deep Learning Toolboxes and RTDS Technologies’ real-time digital simulator with accurate power networks, switching levels of power electronic converters, and a nonlinear battery model.

Research focuses on the efficacy of Multi-Task Autoencoder (MTAE) models in signal classification due to their ability to handle many tasks while improving feature extraction. However, researchers have not thoroughly investigated the study of lung sounds (LSs) for pulmonary disease detection. This paper introduces a new framework that utilizes an MTAE model to detect lung diseases based on LS signals. The model integrates an autoencoder and a supervised classifier, simultaneously optimizing both classification accuracy and signal reconstruction. Furthermore, we propose a hybrid approach that combines an MTAE and a Support Vector Machine (MTAE-SVM) to enhance performance. We evaluated our model using LS signals from a publicly available database from King Abdullah University Hospital. The model attained an accuracy of 89.47% for four classes (normal, pneumonia, asthma, and chronic obstructive pulmonary disease) and 90.22% for three classes (normal, pneumonia, and asthma cases). Using the MTAE-SVM, the accuracy was further improved to 91.49% for four classes and 93.08% for three classes, respectively. The results indicate that the MTAE and MTAE-SVM have a considerable potential for detecting pulmonary diseases from lung sound signals. This could aid in the creation of more user-friendly and effective diagnostic tools.