Search

The effective pricing of electric vehicle (EV) charging by aggregators constitutes a key problem toward the realization of the significant EV flexibility potential in deregulated electricity systems and has been addressed by previous work through bi-level optimization formulations. However, the solution approach adopted in previous work cannot capture the discrete nature of the EV charging/discharging levels. Although reinforcement learning (RL) can tackle this challenge, state-of-the-art RL methods require discretization of state and/or action spaces and thus exhibit limitations in terms of solution optimality and computational requirements. This article proposes a novel deep reinforcement learning (DRL) method to solve the examined EV pricing problem, combining deep deterministic policy gradient (DDPG) principles with a prioritized experience replay (PER) strategy and setting up the problem in multi-dimensional continuous state and action spaces. Case studies demonstrate that the proposed method outperforms state-of-the-art RL methods in terms of both solution optimality and computational requirements and comprehensively analyze the economic impacts of smart-charging and vehicle-to-grid (V2G) flexibility on both aggregators and EV owners. [ABSTRACT FROM AUTHOR]

The increasing penetration of electric vehicles (EV) and fast charging stations (FCS) is tightly coupling the operation of power and transportation systems. In this context, the characterization of the EV flows and charging demand in response to varying traffic conditions and coordinated optimization strategies play a vital role. Previous work on the computation of stochastic traffic user equilibrium (TUE) involve non-linearities in the traffic link and FCS congestion representations and are generally inefficient in dealing with the multi-source uncertainties associated with the operating conditions of the traffic network (TN) and power distribution network (PDN). To address this, this paper proposed a novel deep learning (DL) based surrogate modeling method, leveraging the strength of edge-conditioned convolutional network (ECCN) and deep belief network (DBN). ECCN enables automatic extraction of spatial dependencies, taking into account both node and edge features characterizing the operation of TN. DBN leverages the value of the extracted features and achieves an accurate mapping between the latter to the EV charging demand and EV flows in the TUE, while adaptively generalizing to the multi-dimensional uncertainties. Case studies on three test systems of different scales (including a real-world case involving the matched TN and PDN of Nanjing city) demonstrate that the proposed surrogate model achieves a higher solution accuracy with respect to the state-of-the-art DL-based methods, and exhibits favorable computational performance. Quantitative results also corroborate the benefits brought by the proposed coordinated spatial optimization of EV flows and charging demand on the operation of both TN and PDN. • Coordinated optimization of EV flows and charging demand is investigated. • A novel deep learning based surrogate modeling framework is proposed. • Edge-conditioned convolutional network is employed to extract spatial dependencies. • Deep belief network forecasts the EV charging demand with generalization capability. • Significant techno-economic benefits for the operation of network nexus. [ABSTRACT FROM AUTHOR]

The increasing penetration of small-scale distributed energy resources (DER) has the potential to support cost-efficient energy balancing in emerging electricity systems, but is also fundamentally affecting the conventional operation paradigm of the latter. In this context, innovative market mechanisms need to be devised to better coordinate and provide incentives for DER to utilize their flexibility. Peer-to-Peer (P2P) energy trading has emerged as an alternative approach to facilitate direct trading between consumers and prosumers interacting in an energy collective and fosters more efficient local demand–supply balancing. While previous research has primarily focused on the technical and economic benefits of P2P trading, little effort has been made towards the incorporation of prosumers' heterogeneous characteristics in the P2P trading problem. Here, we address this research gap by classifying the participating prosumers into multiple clusters with regard to their portfolio of DER, and analyzing their trading decisions in a simulated P2P trading platform. The latter employs the mid-market rate (MMR) local pricing mechanism to enable energy trading among prosumers and penalizes the contribution to the system demand peak of each prosumer. We formulate the P2P trading problem as a multi-agent coordination problem and propose a novel multi-agent deep reinforcement learning (MADRL) method to address it. The proposed method is founded on the combination of the multi-agent deep deterministic policy gradient (MADDPG) algorithm and the technique of parameter sharing (PS), which not only enables accelerating the training speed by sharing experiences and learned policies between all agents in each cluster, but also sustains the policies' diversity between multiple clusters. To address the non-stationarity and computational complexity of MADRL as well as persevering the privacy of prosumers, the P2P trading platform acts as a trusted third party which augments the market collective trading information to help training of prosumer agents. Experiments with a large-scale real-world data-set involving 300 residential households demonstrate that the proposed MADRL method exhibits a strong generalization capability in the test data-set and outperforms the state-of-the-art MADRL methods with regard to the system operation cost, demand peak as well as computational time. • The coordinated management of large-scale P2P energy trading is investigated. • The heterogeneity of prosumers with diverse resources' portfolios is accounted. • A novel multi-agent deep reinforcement learning approach is proposed. • The proposed approach achieves significant operating cost and peak demand benefits. • The proposed approach deals effectively with uncertain parameters of the problem. [ABSTRACT FROM AUTHOR]

With the roll-out of smart meters and the increasing prevalence of distributed energy resources (DERs) at the residential level, end-users rely on home energy management systems (HEMSs) that can harness real-time data and employ artificial intelligence techniques to optimally manage the operation of different DERs, which are targeted toward minimizing the end-user's energy bill. In this respect, the performance of the conventional model-based demand response (DR) management approach may deteriorate due to the inaccuracy of the employed DER operating models and the probabilistic modeling of uncertain parameters. To overcome the above drawbacks, this paper develops a novel real-time DR management strategy for a residential household based on the twin delayed deep deterministic policy gradient (TD3) learning approach. This approach is model-free, and thus does not rely on knowledge of the distribution of uncertainties or the operating models and parameters of the DERs. It also enables learning of neural-network-based and fine-grained DR management policies in a multi-dimensional action space by exploiting high-dimensional sensory data that encapsulate the uncertainties associated with the renewable generation, appliances' operating states, utility prices, and outdoor temperature. The proposed method is applied to the energy management problem for a household with a portfolio of the most prominent types of DERs. Case studies involving a real-world scenario are used to validate the superior performance of the proposed method in reducing the household's energy costs while coping with the multi-source uncertainties through comprehensive comparisons with the state-of-the-art deep reinforcement learning (DRL) methods. [ABSTRACT FROM AUTHOR]

The increasing effects of global warming and energy depletion have raised concerns about the pollution caused by traditional oil and fossil energy usage. Distributed energy resources (DERs) have emerged as promising techniques to address these issues. However, the growing proportion of power injection from DERs presents technical challenges such as power quality decline and power flow reversal, which may destabilize the grid. Non-intrusive load monitoring (NILM) is a promising and cost-effective approach that can provide residential power information to improve grid scheduling, dispatching, and optimize residential power consumption behavior. This paper proposes a novel NILM method that can decompose the power of both household appliances and DERs integrated with residential houses. The proposed approach employs a data segment method that utilizes adaptive window lengths to identify the behavior of DERs based on the operating periods and characteristics of different equipment. A non-intrusive load monitoring method is proposed based on the ResNet-seq2seq framework to address the problem of model degradation and gradient vanishing. Post-processing techniques are applied to the output results using a feature list of the target equipment. Experiments are conducted on a dataset organized based on the REDD and Pecan Street datasets to demonstrate the accuracy and effectiveness of the proposed method. The results show that compared to the start-of-the-art CNN-seq2seq model, the proposed method achieved 54.95%, 69.18%, 56.95%, and 84.98% reduction in MAE for PV, EV, refrigerators, and microwaves, respectively. • A novel ResNet-seq2seq-based NILM approach is proposed for load disaggregation of residential houses integrating DERs. • Wavelet denoising and adaptive window length data segments are used for data pre-processing and accurate disaggregation. • Residual connections are employed to solve model degradation and vanishing gradient issues as network depth increases. • A post-processing technique is applied to modify the output of the network and improve disaggregation accuracy. [ABSTRACT FROM AUTHOR]

• A two-stage localisation framework for forced oscillations is proposed. • Graphical representation is generated to characterise forced oscillations. • Knowledge of image recognition is transferred to the localisation problem. • It has high accuracy, good anti-noise performance and robustness. Accurately locating forced oscillations (FOs) disturbance source in a large-scale power system is a challenging task. In this paper, a localisation method featuring system-level and area-level localisation is proposed. Compared with traditional localisation methods which process the FOs signals only in control centre phasor data concentrator (PDC), the proposed method can not only relieve the data communication pressure, but also meets the data privacy and confidentiality requirements of utility companies. Besides, with the two-stage deep transfer learning, high accuracy of localisation can be realised with far less training. Firstly, by adopting principle component analysis (PCA) to extract the most representative information at system-level and smooth pseudo Wigner-Ville distribution (SPWVD) to characterize FOs signals, graphical representations of FOs at system-level and area-level are obtained respectively, transforming the localisation problem to image recognition problem. Subsequently, a two-stage deep transfer learning (DTL) algorithm is developed to locate FOs disturbance source. The first stage involves repurposing the learnt features from a pre-trained deep convolutional neural network (previously trained for universal image recognition) to improve the learning of system-level localisation. In a similar logic, the second stage entails transferring the knowledge acquired from the first stage is exploited to aid area-level localisation learning. Case studies carried out on the WECC 179-bus system demonstrate that the proposed method achieves a significantly higher accuracy in the presence of measurement noise, topology variation and load disturbances of the system with respect to the traditional machine learning method. The proposed method also exhibits a more favorable computational performance and learning efficiency due to the employment of the tactfully designed two-stage DTL approach. [ABSTRACT FROM AUTHOR]