• Energy Research

In response to electricity markets dominated by wind energy production, and thereby varying electricity prices, this research aims at examining intensives for investments in integrated renewable energy power systems. To do so, strategies have been presented and discussed using optimization methodologies for a power system consisting of wind turbines, electrolysers, and fuel cells. Consequently, intensives for investments in such power systems are introduced by optimizing the return on economic investments of wind farms in markets with great daily variations in electricity prices. The findings presented in this research can help inform researchers, policy makers, and the energy industry in the transition towards implementation of renewable energy systems.

With the application of advanced information technology for the integration of electricity and natural gas systems, formulating an excellent energy conversion and management strategy has become an effective method to achieve established goals. Differing from previous works, this paper proposes a peak load shifting model to smooth the net load curve of an integrated electricity and natural gas system by coordinating the operations of the power-to-gas unit and generators. Moreover, the study aims to achieve multi-objective optimization while considering the economy of the system. A dynamic energy conversion and management strategy is proposed, which coordinates both the economic cost target and the peak load shifting target by adjusting an economic coefficient. To illustrate the complex energy conversion process, deep reinforcement learning is used to formulate the dynamic energy conversion and management problem as a discrete Markov decision process, and a deep deterministic policy gradient is adopted to solve the decision-making problem. By using the deep reinforcement learning method, the system operator can adaptively determine the conversion ratio of wind power, power-to-gas and gas turbine operations, and generator output through an online process, where the flexibility of wind power generation, wholesale gas price, and the uncertainties of energy demand are considered. Simulation results show that the proposed algorithm can increase the profit of the system operator, reduce wind power curtailment, and smooth the net load curves effectively in real time.

Abstract Water abandonment in hydroelectricity production is a major challenge that can be solved with an increase in electricity demand. China as a country with huge hydropower installation is faced with the problem of underutilizing the hydropower potential due to inadequate electricity demand and transmission facility. In this study, we investigate the potential of hybridizing hydrogen production with hydropower stations in Southwestern China. We found that the integration of hydrogen production with hydropower stations will help reduce the country's CO2 emissions and as much as 1.18% reduction in China emission can be achieved adopting this methodology. In a hydropower station of 750 MW capacity, about 3.142 × 108 kg of hydrogen could have been produced from the abandoned water in 2019. This will also result in 351, 734, 330.9 kgCO2/yr emission reduction. We also developed a model to determine the optimized hydrogen installed capacity based on different parameters. Based on 2019 data, the CO2 emission of China will be reduced by 0.127% with the production of hydrogen from the excess electricity of a 750 MW hydropower station (Case study A) in Sichuan Province.

Due to the increasing penetration level of wind power in power system, wind turbines can provide less frequency support than conventional generators due to their small rotor mass. This makes the power system with low inertia and hence cause frequency problem. This paper has proposed an integrated control strategy for participate in primary frequency control by doubly fed induction generation (DFIG). According to the reserve capacity required for primary frequency control, a de-loading control method is also proposed in this paper to resolve the issue of inertia control and primary frequency control. Based on the analysis method of the frequency control characteristics of DFIGs, the frequency control strategy can adjust the static frequency difference coefficient and it is proposed by improved variable pitch control method. Furthermore, the virtual inertia control and the primary frequency control can be integrated a control strategy of DFIGs. The simulation results show that the DFIGs can provide an effective inertia support to reduce the system frequency changing rate in the inertia process and improve the static frequency stability of power system by the static frequency characteristics.

This paper proposes a novel data-driven self-tuning additional sliding mode controller for power system transient voltage stability enhancement considering wind integration. We first develop a new additional fractional-order sliding mode controller (FOSMC) for the static var compensator (SVC). The tuning of FOSMC parameter settings is then reformulated as a Markov decision process (MDP) and solved by the deep reinforcement learning (DRL) algorithm. In addition, a data-driven estimation method is proposed for the identification of an equivalent transfer function that is further used to calculate reward during the training process, yielding the model-free training. After that, the well-trained agent allows us to tune the controller parameters considering system uncertainties and achieve robustness against various operating conditions. Comparative results with other state-of-art methods demonstrate that the proposed method can effectively suppress the chattering issue of the sliding mode controller and ensure transient voltage stability under different operating conditions.

This paper proposes attention enabled multi-agent deep reinforcement learning (MADRL) framework for active distribution network decentralized Volt-VAR control. Using the unsupervised clustering, the whole distribution system can be decomposed into several sub-networks according to the voltage and reactive power sensitivity relationships. Then, the distributed control problem of each sub-network is modeled as Markov games and solved by the improved MADRL algorithm, where each sub-network is modeled as an adaptive agent. An attention mechanism is developed to help each agent focus on specific information that is mostly related to the reward. All agents are centrally trained offline to learn the optimal coordinated Volt-VAR control strategy and executed in a decentralized manner to make online decisions with only local information. Compared with other distributed control approaches, the proposed method can effectively deal with uncertainties, achieve fast decision makings, and significantly reduce the communication requirements. Comparison results with model-based and other data-driven methods on IEEE 33-bus and 123-bus systems demonstrate the benefits of the proposed approach.

The uncertainties of charging behavior of electric vehicle (EV) owners have a negative impact on the loss of life (LOL) of distribution transformer. This article proposes a decentralized EV charging framework for optimization of the LOL of distribution transformer considering the dissatisfactions of EV owners. Specifically, long-short-term memory (LSTM) neural network is first utilized to capture the uncertainties caused by the load demand and electricity price. After that, each EV is modeled as an intelligent agent and a multiagent deep reinforcement learning approach is applied to solve the coordinated charging problem based on the forecasting information by the LSTM network. All the agents are trained in a centralized manner to develop coordinated control strategies while informing decisions based on local information when finishing the training process. The proposed approach can achieve coordinated charging management of EVs based on local information, which helps preserve the privacy of EV owners, reduce the cost induced by the deployment of communication devices, and avoid single-point failure. In addition, the parameter space noise and deep dense architecture in reinforcement learning are introduced to overcome premature convergence, training instability, and inefficiency due to the large action space of multiagent scenario. Comparative tests are carried out among several benchmarks utilizing real-world data to illustrate the effectiveness of the proposed approach.

As the scale of onshore wind farms are increasing, the influence of wake behavior on power production becomes increasingly significant. Wind turbines sittings in onshore wind farms should take terrain into consideration including height change and slope curvature. However, optimized wind turbine (WT) placement for onshore wind farms considering both topographic amplitude and wake interaction is realistic. In this paper, an approach for optimized placement of onshore wind farms considering the topography as well as the wake effect is proposed. Based on minimizing the levelized production cost (LPC), the placement of WTs was optimized considering topography and the effect of this on WTs interactions. The results indicated that the proposed method was effective for finding the optimized layout for uneven onshore wind farms. The optimization method is applicable for optimized placement of onshore wind farms and can be extended to different topographic conditions.