Search

In a real-world scenario of load forecasting, it is crucial to determine the energy consumption in electrical networks. The energy consumption data exhibit high variability between historical data and newly arriving data streams. To keep the forecasting models updated with the current trends, it is important to fine-tune the models in a timely manner. This article proposes a reliable inductive transfer learning (ITL) method, to use the knowledge from existing deep learning (DL) load forecasting models, to innovatively develop highly accurate ITL models at a large number of other distribution nodes reducing model training time. The outlier-insensitive clustering-based technique is adopted to group similar distribution nodes into clusters. ITL is considered in the setting of homogeneous inductive transfer. To solve overfitting that exists with ITL, a novel weight regularized optimization approach is implemented. The proposed novel cross-model methodology is evaluated on a real-world case study of 1000 distribution nodes of an electrical grid for one-day ahead hourly forecasting. Experimental results demonstrate that overfitting and negative learning in ITL can be avoided by the dissociated weight regularization (DWR) optimizer and that the proposed methodology delivers a reduction in training time by almost 85.6% and has no noticeable accuracy losses. [ABSTRACT FROM AUTHOR]

The energy Internet (EI) equipment may face threats that attackers poison federated learning (FL) models to disturb electricity load forecasting. To mitigate this vulnerability, it is important to study load forecasting disturbance approaches. This article proposes a side-channel analysis (SCA)-based disturbance approach. First, we design an FL SCA scheme to extract power information from the FL chip running forecasting model. Second, we propose an FL data speculation method using an optimized convolutional neural network trained with SCA information. Third, we design a label-flipping-based poisoning scheme with speculated data characteristics for load forecasting disturbance. Experimental results show attackers can successfully poison and disturb FL-based load forecasting. The average accuracy of EI load data speculation is 99.8%. This work is the first to study EI load forecasting disturbance from an SCA perspective. [ABSTRACT FROM AUTHOR]

Electricity distribution companies are in-charge-of supplying stable electric energy to urban customers over a wide geographical area. In these distribution networks, large number of Distribution-Transformers (DTs) are installed on towers or in electric-rooms with the ratings below 1000 kVA, which are the most important assets of distribution companies. Periodic maintenance and test services are not economically feasible for all in-operation transformers. Hence, estimating the remaining useful life and scheduling the maintenance times of DTs are of great technical and economic importance. An analytical approach is presented in this paper for predicting lifetime of DTs based on degree of polymerization. An appropriate Machine-Learning (ML) algorithm trained by measured data is selected for forecasting driven load according to the ambient temperature and humidity and day-hour schedules. The relation of Hot-Spot-Temperature in IEC60076∼7 is modified based on the measured data for accounting unbalanced-loading. Neutral-current is predicted using a trained ML-model. Another approach is proposed for determining times of oil filtering services aiming to minimize the maintenance cost and maximizing the remaining life. The proposed lifetime estimation method is validated by comparing its results with the damaged DT data of a sample network chosen as the case studies for this research, i.e., Bushehr province, Iran. [ABSTRACT FROM AUTHOR]

Power system inertia is declining and is increasingly variable and uncertain in regions where the penetration of non-synchronous generation and interconnectors is growing. This presents a challenge to power system operators who must take appropriate actions to ensure the stability and security of power systems relying on short-term forecasts of the system’s inertial response. Existing models to forecast inertia fail to quantify uncertainty, which may prevent their utilization given the risk aversion of the system operators when handling stability issues. This paper is the first to develop a model to produce calibrated, data-driven probabilistic forecasts of the inertia contribution of transmission-connected synchronous generators. The model provides a necessary tool for system operators to quantify forecast uncertainty, allowing them to manage the risk of frequency instability cost-effectively. The paper demonstrates that the assumption of a Gaussian distribution of uncertainty applied in existing models is not acceptable to accurately forecast the inertial response and provides a satisfactory forecast model by combining non-parametric density forecasting with parametric tail distributions. Moreover, the paper shows that satisfactory predictive performance can only be achieved by adopting a rolling horizon forecast approach to deal with the rapidly changing characteristics of the inertial response in power systems. [ABSTRACT FROM AUTHOR]

Long-term electricity consumption forecasting has been a hot issue due to its importance in power systems planning. Recently, system dynamics (SD) have been considered a potential method for electricity consumption forecasting. However, traditional SD methods are failed in forecasting future power systems toward low-carbon, due to the complex influencing factors and the simple SD model structure. In this study, a novel method combining SD and the ImPACT equation is proposed to solve the above problem. First, a new influencing factor system including emerging factors is established to consider the development of future power system. Subsequently, the ImPACT equation is employed to decompose the electricity consumption into four subsystems. By doing this, the linkage between factors and electricity consumption can be better revealed, which helps to improve the forecasting accuracy. Next, models for quantifying the correlation between subsystems and factors are established using SD. Finally, the proposed method is verified by forecasting the power consumption of China in the future year 2030 to 2050. Furthermore, various scenarios are set to analyze the changing trend of future power consumption by different influencing factors. The results of this article can provide important basic information for power supply and power grid planning. [ABSTRACT FROM AUTHOR]

Although the idea of information granulation has been shown to be a research craze in short-term time series forecasting, it is still urgent to develop a granular framework so that information granulation can characterize the trend distribution of data to the significant extent under a common concept of time. This article puts forward a novel granulation algorithm involving two-stage partitioning scheme so that information granule established there exhibits well-articulated semantics at the time level, while at the same time, it gives full consideration to the varying patterns of data. On this basis, a new association rule based on this type of information granules is presented. Unlike most fuzzy association rules, the proposed association rules can extract and derive the correlations between two collections of trend features corresponding to the past and future, which is in accord with human's reasoning. Keeping in mind that the prediction process should center on essential rules while freeing from the interference of irrelevant rules, which contributes to a reliable prediction result, thus, a rule selection algorithm is involved so as one makes sure the accuracy and interpretability of the forecasting results. The design of short-term forecasting model based on fuzzy inference system is implemented, where the concept of granulation eliminates the commonly used alternative, i.e., the recursive iterations of one-step prediction. Experimental results have verified the effectiveness of the proposed model. [ABSTRACT FROM AUTHOR]

This letter proposes a transferrable model-agnostic meta-learning (T-MAML) approach for short-term load forecasting for single households. The proposed approach enables multiple households to collaboratively train a generic artificial neural network (ANN) model. The generic ANN model is then further trained at each target household node for the STLF purpose. The proposed T-MAML based STLF approach is featured by: (1) significant reduction of computation and communication costs on the household side; and (2) superior STLF performance, especially when there is limited load data for training in a target household. Experiments based on a real Australian residential dataset are conducted to validate the effectiveness of the proposed approach. [ABSTRACT FROM AUTHOR]

As an important application in the power system operation and electricity market clearing, the network-constrained unit commitment (NCUC) problem is usually executed by Independent System Operators (ISO) in an open-looped predict-then-optimize (O-PO) process, in which an upstream prediction (e.g., on renewable energy sources (RES) and loads) and a downstream NCUC are executed in a queue. However, in the O-PO framework, a statistically more accurate prediction may not necessarily lead to a higher NCUC economics against actual RES and load realizations. To this end, this paper presents a closed-loop predict-and-optimize (C-PO) framework for improving the NCUC economics. Specifically, the C-PO leverages structures (i.e., constraints and objective) of the NCUC model and relevant feature data to train a cost-oriented RES prediction model, in which the prediction quality is evaluated via the induced NCUC cost instead of the statistical forecast errors. Therefore, the loop between the prediction and the optimization is closed to deliver a cost-oriented RES power prediction for NCUC optimization. Lagrangian relaxation is adopted to accelerate the training process, making the C-PO applicable for real-world systems. Case studies on an IEEE RTS 24-bus system and an ISO-scale 5655-bus system with real-world data show that the proposed C-PO can effectively improve the NCUC economics as compared to the traditional O-PO. [ABSTRACT FROM AUTHOR]

As an efficient tool for uncertainty quantification of renewable energy forecasting, prediction intervals (PIs) provide essential prognosis to power system operator. Merely improving the statistical quality of PIs with respect to calibration and sharpness cannot always contribute to the operational value for specific decision-making issue. In this paper, the cost-oriented prediction intervals are firstly proposed to achieve the joint improvement of forecasting quality and decision performance. In order to bridge the gap between forecasting and decision, a novel cost-oriented machine learning (COML) framework is established, which unifies nonparametric renewable power PI construction and decision-making. Formulated as a bilevel programming model, the COML minimizes the operational costs of decision-making process by adaptively adjusting the quantile proportion pair of PIs resulting from extreme learning machine based quantile regression. The hierarchical optimization model of the COML is equivalently simplified as a single level nonlinear programming problem. Then an enhanced branch-and-contract algorithm with innovative bounds contraction strategy is devised to efficiently capture the optimum of the single level problem with bilinear nonconvexity. Numerical experiments based on actual wind farm data simulate the online forecasting and decision process for wind power offering. Comprehensive comparisons verify the substantial superiority of the proposed COML methodology in terms of forecasting quality, operational value, as well as computational efficiency for practical application. [ABSTRACT FROM AUTHOR]

In recent years, a large variety of short-term load forecasting methodologies have been proposed, but a common drawback of them is that require a large amount of historical data to train the model. However, in reality, the cases in limited historical data would be more frequently occurred, for example, a new house or substation is built. To tackle this issue, a combining forecasting method of deep residual neural network (ResNet)-based transfer learning for residential loads is proposed in this article. To improve the accuracy and reliability of transfer learning, a novel deep ResNet with dual skip connections (ResNet-DSC) is proposed as the base model. Then with sparse training data, a Bayesian probability weighted averaging (approach is proposed, to address the model combination parameters estimation problem. In addition, probability density function forecasting is also delivered by the method. For demonstration, a series of experiments, with actual residential load data, was performed. With the comparison with other transfer learning and nontransfer learning approaches, the effectiveness and improvement of the proposed method have been validated. [ABSTRACT FROM AUTHOR]

The reliable operation of a power distribution system relies on a good prior knowledge of its topology and its system state. Although crucial, due to the lack of direct monitoring devices on the switch statuses, the topology information is often unavailable or outdated for the distribution system operators for real-time applications. Apart from the limited observability of the power distribution system, other challenges are the nonlinearity of the model, the complicated, unbalanced structure of the distribution system, and the scale of the system. To overcome the above challenges, this paper proposes a Bayesian-inference framework that allows us to simultaneously estimate the topology and the state of a three-phase, unbalanced power distribution system. Specifically, by using the very limited number of measurements available that are associated with the forecast load data, we efficiently recover the full Bayesian posterior distributions of the system topology under both normal and outage operation conditions. This is performed through an adaptive importance sampling procedure that greatly alleviates the computational burden of the traditional Monte-Carlo (MC)-sampling-based approach while maintaining a good estimation accuracy. The simulations conducted on the IEEE 123-bus test system and an unbalanced 1282-bus system reveal the excellent performances of the proposed method. [ABSTRACT FROM AUTHOR]

The stochastic cloud cover on photovoltaic (PV) panels affects the solar power outputs, producing high instability in the integrated power systems. It is an effective approach to track the cloud motion during short-term PV power forecasting based on data sources of satellite images. However, since temporal variations of these images are noisy and non-stationary, pixel-sensitive prediction methods are critically needed in order to seek a balance between the forecast precision and the huge computation burden due to a large image size. Hence, a graphical learning framework is proposed in this study for intra-hour PV power prediction. By simulating the cloud motion using bi-directional extrapolation, a directed graph is generated representing the pixel values from multiple frames of historical images. The nodes and edges in the graph denote the shapes and motion directions of the regions of interest (ROIs) in satellite images. A spatial-temporal graph neural network (GNN) is then proposed to deal with the graph. Comparing with conventional deep-learning-based models, GNN is more flexible for varying sizes of input, in order to be able to handle dynamic ROIs. Referring to the comparative studies, the proposed method greatly reduces the redundancy of image inputs without sacrificing the visual scope, and slightly improves the prediction accuracy. [ABSTRACT FROM AUTHOR]

Uncertainties in renewable generation make accurate load forecast essential for reliable power system operation. This paper considers energy imbalance market (EIM), where market players are allowed to procure energy ahead of time and trade the mismatch due to forecast error and strategic behaviors. The ISO sets the trading prices according to the market conditions, and pursues various system-level objectives. We first identify the power-law relationship between data volume and forecast accuracy, which enables the formulation of forecast cost model. Then, we cast the interactions in the EIM in the Stackelberg game framework with the ISO acting as the leader. We offer explicit subgame perfect equilibrium among the players in EIM, and derive the sufficient condition for the existence of unique equilibrium. Then, we show that this equilibrium, if exists, supports the maximal social welfare and under certain conditions, minimizes the total mismatch. We further examine the local and global impacts of the forecast errors under mild conditions, together with robustness analysis. Such analysis provides mechanism design guidelines for the ISO to enable the data sharing and forecast method sharing among market players in the EIM. Numerical studies further examine the effectiveness, robustness and sensitivity of the subgame. [ABSTRACT FROM AUTHOR]

The solutions recommended by international roadmaps and technical reports on smart metering refer to interval metering with time resolutions higher than 15 min to 1 h as traditionally used. Based on the characteristics of the users’ power patterns in distribution networks, this article shows that in practical cases the resolutions of the traditional metering are not sufficient to assess peak power and network losses effectively. Effective interval metering solutions should have resolutions of one minute or less. Moreover, this article shows the advantages of assessing the average power peak (amplitude and duration) and estimating the network losses through innovative solutions beyond interval metering, based on event-driven energy metering. The use of EDM significantly enhances the quality of pattern representation and reduces the amount of data required with respect to high-resolution interval metering. Based on the Pareto analysis of conflicting objectives, a novel procedure to set up the EDM thresholds is presented. The applications shown use real data and refer to a single user, some users connected to a distribution network feeder, and many users connected to a large distribution system. The EDM benefits are quantified using specific indicators that consider energy losses and peak power. [ABSTRACT FROM AUTHOR]

In structural health monitoring (SHM), measuring and evaluating structural dynamic responses are critical for safety management of civil infrastructures. Particularly, online forecasting of the structural responses under extreme external loading conditions (e.g., earthquakes) takes a significant role in SHM to provide early warning and ensure safe operation. In practice, complex causality and intrinsic interactions between seismic excitation and structural response make it challenging to establish a reliable predictive scheme. The present paper proposes a novel deep recurrent neural network (RNN) model implemented in the architecture of a time-series attention-based RNN encoder–decoder (TSA-RNN-ED), for predictive analysis of structural responses under seismic excitation. In the proposed data-driven model, upcoming sequential responses are predicted through sequence-to-sequence learning from historical multivariate time-series signals. A time-series attention mechanism is proposed to exploit the heterogeneous, but directly related, hidden features between the seismic loads and the corresponding structural responses. The proposed architecture can reliably regress excitation-response interactions to predict dynamic responses subjected to future earthquakes while satisfying the need of real-time forecasting for on-site practical implementation. This article systematically evaluates the proposed model by using two real-world structural cases: 1) the tallest building in China, the Shanghai Tower and 2) a woodframe classroom on a shake table at the University of British Columbia in Vancouver, Canada. The experimental results demonstrate the accurate and efficient performance of the proposed methodology in forecasting the seismic responses of the structures under investigation. [ABSTRACT FROM AUTHOR]

Friction, generated in the contact patch between tires and runway, has long been recognized as a key issue to a safe landing. Moreover, soft sensing technology is also considered to be one of the effective means to improve measurement accuracy and reduce measurement cost of friction coefficient. This article presents a study on the prediction methods of runway friction coefficient based on multivariable coupling. Firstly, the relationships between friction coefficient and tread frictional stress under different working conditions are investigated using finite element analysis; meanwhile, the modeling dataset and test datasets for multivariate coupling are established. Secondly, regression analysis and BP neural network (BPNN) are used to develop prediction models between independent variables (inflation pressure, load, velocity, slip, and frictional stress) and dependent variable (friction coefficient). Finally, the prediction models are evaluated based on internal and external verification. The results demonstrate the superior performance of both prediction models. After further comparison, it is concluded that the performance of BPNN is even better combining the error range and model evaluation indicators including MAE, RMSE, and ${R}^{2}$. [ABSTRACT FROM AUTHOR]

The failure of ac relay contacts will cause electrical control system fault, which will bring certain economic losses. Therefore, accurate prediction and evaluation of the electrical life of the relay have certain research value and practical significance. The measurement of the contact resistance of the relay was performed using a self-designed experimental platform. An electrical life prediction method of ac relay was proposed. First, the measured contact resistance signal was decomposed using empirical mode decomposition (EMD), and several sub-sequence components and residual components were obtained. Second, prediction of these components was conducted using a long short-term memory neural network (LSTM), and the prediction value was reconstructed. Third, the group method of data handling (GMDH) algorithm and Harris Hawk optimization (HHO) algorithm were used to optimize LSTM. Finally, contact pressure was calculated using the predicted contact resistance. The electrical life of the relay was predicted according to a contact-pressure-based electrical life criterion. Compared with other models, the prediction accuracy of the contact resistance is higher. The electrical life of the relay can be predicted effectively using the prediction results of both contact resistance and contact pressure. In practical application, a pressure sensor can be installed in the relay, and the electrical life of the contact can be predicted by the measured contact pressure. [ABSTRACT FROM AUTHOR]

The increasing penetration of solar generation into power grids has promoted the need for accurate and reliable short-term solar irradiance forecasting. This article introduces a novel multibranch attentive gated recurrent residual network (ResAttGRU) consisting of multiple branches of residual networks, gated recurrent units (GRUs), and the attention mechanism. The proposed multibranch ResAttGRU is capable of modeling data at various resolutions, extracting hierarchical features, and capturing short- and long-term dependencies. Moreover, this network also presents a strong multitimescale representative within the proposed architecture, while GRUs can exploit temporal information at less computational cost than the popular long short-term memory (LSTM). The novelty of the proposed architecture is to employ multiple convolutional-based branches with different filter lengths to learn multitimescale features jointly, accelerate the learning process, and reduce overfitting by leveraging shared representations as the auxiliary information. This study also compares the multibranch ResAttGRU networks with state-of-the-art deep learning methods using 18 years of NSRDB data at 12 solar sites. Finally, the proposed multibranch ResAttGRU requires 7.1% fewer parameters than multibranch residual LSTM while achieving similar average RMSE, MAE, and R-squared values. [ABSTRACT FROM AUTHOR]

In this article, we propose a distributed demand-side management (DSM) approach for smart grids taking into account uncertainty in wind power forecasting. The smart grid model comprehends traditional users as well as active users (prosumers). Through a rolling-horizon approach, prosumers participate in a DSM program, aiming at minimizing their cost in the presence of uncertain wind power generation by a game theory approach. We assume that each user selfishly formulates its grid optimization problem as a noncooperative game. The core challenge in this article is defining an approach to cope with the uncertainty in wind power availability. We tackle this issue from two different sides: by employing the expected value to define a deterministic counterpart for the problem and by adopting a stochastic approximated framework. In the latter case, we employ the sample average approximation (SAA) technique, whose results are based on a probability density function (PDF) for the wind speed forecasts. We improve the PDF by using historical wind speed data, and by employing a control index that takes into account the weather condition stability. Numerical simulations on a real data set show that the proposed stochastic strategy generates lower individual costs compared to the standard expected value approach. [ABSTRACT FROM AUTHOR]

Electricity price is a crucial element for market players to maximize their profits. In this context, the forecast of the hour-ahead, day-ahead, and week-ahead electricity prices plays a crucial role. The more accurate the prediction is, the lower the market risk is. In this article, several machine learning algorithms (support vector machine, Gaussian processes regression, regression trees, and multilayer perceptron) with different structures have been adopted to forecast Italian wholesale electricity prices. Considering different time horizons (hourly, daily, and weekly), their performances have been compared through several performance metrics, including mean absolute error, R-index, mean absolute percentage error, and the number of anomalies in which the forecast error passes a threshold. The investigation reveals that, in general, SVM and tree-based models outperform other models at different time horizons. [ABSTRACT FROM AUTHOR]

In this research work, a deep learning (DL)-based method for the fast and accurate analysis of current harmonics of electric arc furnaces (EAF) is proposed. For such a system, a large amount of EAF current data is required for the training phase of the DL-based structure, which is not only a thorny but also an expensive procedure. Hence, the second focus of this research work is to gain the ability to generate EAF currents with realistic harmonic contents based on a much smaller amount of field data of EAF currents. For this purpose, EAF current data, recorded at a transformer substation supplying an EAF plant during a tap-to-tap time of the EAF operation, are examined in terms of harmonic component amplitudes and phases. Then, a significantly larger amount of EAF current data is regenerated based on the statistics of current harmonics mimicking the real EAF behavior and this synthetic data are used to train the DL-based harmonic estimator. This estimator is able to estimate both amplitudes and phases of the harmonics without computing any time- or frequency-domain features during the estimation process. Hence, the outcomes of this research work are twofold: First, detailed analysis of the EAF current harmonic behavior is achieved, which reveals the operation principles of the EAF. Second, a DL-based harmonic estimator is trained, which is able to output the amplitude and phase estimations directly out of waveform samples without any feature extraction. The proposed system aims to serve the needs of active power filters of the EAF installations in the electricity system, since it has been shown that fast and accurate harmonic amplitude and phase estimations are obtained. [ABSTRACT FROM AUTHOR]

Load forecasting is one of the most important and studied topics in modern power systems. Most of the existing research on day-ahead load forecasting try to build a good model to improve the forecasting accuracy. The forecasted load is then used as the input to generation scheduling with the ultimate goal of minimizing the cost of generation schedules. However, existing day-ahead load forecasting models do not consider this ultimate goal at the training/forecasting stage. This paper proposes a task-based day-ahead load forecasting model labeled as LfEdNet that combines two individual layers in one model, including a load forecasting layer based on deep neural network (Lf layer) and a day-ahead stochastic economic dispatch (SED) layer (Ed layer). The training of LfEdNet aims to minimize the cost of the day-ahead SED in the Ed layer by updating the parameters of the Lf layer. Sequential quadratic programming (SQP) is used to solve the day-ahead SED in the Ed layer. The test results demonstrate that the forecasted results produced by LfEdNet can lead to lower cost of day-ahead SED at the expense of slight reduction in forecasting accuracy. [ABSTRACT FROM AUTHOR]

Uncertainty modeling of renewable energy sources, load demand, electricity price, etc. create a high volume of data in smart grids. Accordingly, in this article, a precise forecasting method based on a deep learning concept with microclustering (MC) task is presented. The MC method is structured based on hybrid unsupervised and supervised clustering tasks by K-means and Gaussian support vector machine, respectively. In the proposed method, the input data sequence is clustered by the MC task, and then the forecasting process is employed. By applying the MC, input data in each hour are categorized into different groups, and a distinctive forecasting unit is allocated to each one. In this way, more clusters and forecasting networks are earmarked for the hours with higher fluctuation rates. The bi-directional long short-term memory (B-LSTM), which is one of the newest recurrent artificial neural networks, is proposed as the forecasting unit. The B-LSTM has bidirectional memory—feedforward and feedback loops—that helps us to investigate both previous and future hidden layers data. The optimal number of clusters in each hour is determined based on the Davies–Bouldin index. To evaluate the performance of the proposed method, in this study, three forecasting tasks including the wind speed, load demand, and electricity price are studied in different periods using the Ontario province, Canada, data set. The results are compared with other benchmarking methods to verify the robustness and effectiveness of the proposed method. In fact, the proposed method, which is equipped with the MC technique and B-LSTM networks, significantly promotes the forecasting results, especially in spike points. [ABSTRACT FROM AUTHOR]

The coronavirus disease 2019 (COVID-19) pandemic has urged many governments in the world to enforce a strict lockdown where all nonessential businesses are closed and citizens are ordered to stay at home. One of the consequences of this policy is a significant change in electricity consumption patterns. Since load forecasting models rely on calendar or meteorological information and are trained on historical data, they fail to capture the significant break caused by the lockdown and have exhibited poor performances since the beginning of the pandemic. In this paper we introduce two methods to adapt generalized additive models, alleviating the aforementioned issue. Using Kalman filters and fine-tuning allows to adapt quickly to new electricity consumption patterns without requiring exogenous information. The proposed methods are applied to forecast the electricity demand during the French lockdown period, where they demonstrate their ability to significantly reduce prediction errors compared to traditional models. Finally, expert aggregation is used to leverage the specificities of each predictions and enhance results even further. [ABSTRACT FROM AUTHOR]

Wind power producers participating in day-ahead electricity markets are compelled to pay imbalance costs if they do not generate the same amount of power as they had bid for. These imbalance costs comprise a significant proportion of their income. To reduce the risk of such financial losses, this paper employs the idea of trading in a separate prediction market, as a hedging method. In prediction markets, participants trade shares associated with a certain outcome of an event. We propose that the wind power producers might participate in a prediction market to trade the future value of the wind power and by taking an opposite position in comparison to the electricity market, the imbalance costs will be offset through payouts in the prediction market. Wind power is modelled as a stochastic variable and an optimal trading strategy is developed where the trading volume in the prediction market is analytically derived and formulated by minimising the maximum possible loss and pricing of shares is determined via indifference utility condition. The results suggest that the proposed method limits the loss values and improves the risk measures. [ABSTRACT FROM AUTHOR]

Smart micro-grid (SMG) is a growing segment of the modern power grid. Besides the benefits brought in terms of improving power reliability, power quality, security, and sustainability, and promoting competitiveness in a new deregulated electricity market, SMG is still in the early commercial stage. This article introduces the SMG con-cept-based forecasting mechanisms. It highlights the role of the power production and profit forecasting methods to drive the SMG deployment with the liberalized electricity market concept. Finally, some open issues in SMG deployment such as standardization and regulations are presented as key drivers for SMG success. [ABSTRACT FROM AUTHOR]

Adequate reserves are urgently needed to hedge against wind power forecasting uncertainties in power systems. Traditional reserve quantification sequentially acquires statistical features of wind power and then determines reserve amounts. This paper establishes a novel integrated probabilistic forecasting and decision (IPFD) methodology to simultaneously optimize the wind power prediction intervals (PIs) and probabilistic reserve quantification. Upward and downward reserve quantities are defined to cover the wind power forecasting uncertainties within the PIs. A cost function evaluating the reserve provision payment and deficit penalty is elaborated to realize cost-benefit trade-offs of reserve decision. Nonparametric wind power PIs are constructed based on extreme learning machine, which minimizes the reserve cost function subject to eligible target coverage probability constraint. The confidence level and quantile proportions associated with wind power PIs can be jointly tuned to reduce the operational cost of reserves. Benefited from extreme learning machine, the IPFD model is reformulated as a mixed integer linear programming problem. A feasible region tightening strategy that shrinks the large constant coefficients and eliminates the redundant binary variables is proposed to accelerate model training. Numerical experiments based on actual wind power data demonstrate the remarkable cost-effective advantages of the IPFD based reserve quantification, as well as the high computational efficiency for online application. [ABSTRACT FROM AUTHOR]

A prior knowledge of residential load demand is critical for power system operations at the distribution level, such as economic dispatch, demand response and energy storage schedule. However, as residential customers perform more casual and active consumption behaviors, prediction of such highly volatile loads can be much harder. Owing to the development of sensor technology, micrometeorological data can be sampled with a high geographic resolution. Those data that represent the weather condition on the land surface show a strong relationship to the residential load evidently, whereas it remains unsolved on how to fully utilize those great number of datasets. This paper proposes a day-ahead probabilistic residential load forecasting method based on a novel deep learning model, named convolutional neural network with squeeze-and-excitation modules (CNN-SE), and micrometeorological data. The model can employ multi-channel input data with dissimilar weights, suitable for analyzing massive relevant input factors. A feature extraction method is adopted for customer consumption pattern based on sparse auto-encoder (SAE), which can help correct probabilistic forecasting results. A case study that covers 8 residential communities and 18 micrometeorological sites is conducted to validate the feasibility and accuracy of the proposed hybrid method. [ABSTRACT FROM AUTHOR]

Load forecasting is of crucial importance for operations of electric power systems. In recent years, deep learning based methods are emerging for load forecasting because their strong nonlinear approximation capabilities can provide more forecasting precision than conventional statistical methods. However, they usually suffer from some problems, e.g., the gradient vanishment and over-fitting. In order to address these problems, an unshared convolution based deep learning model with densely connected network is proposed. In this model, the backbone is the unshared convolutional neural network and a densely connected structure is adopted, which could alleviate the gradient vanishment. What is more, we use a regularization method named clipped $L_2$ -norm to overcome over-fitting, and design a trend decomposition strategy to address the possible distribution differences between the training and validation data. Finally, we conduct five case studies to verify the outperformance of our proposed deep learning model for deterministic and interval load forecasting. Two high-voltage and an medium-voltage real load datasets from Australia, Germany and America are used for model training and validation, respectively. Results show that the proposed model can achieve higher load forecasting accuracy, compared with other existing methods including the popular conventional methods such as naive forecast and generalized additive model, and deep learning methods, e.g., long short-term memory network, convolutional neural network, fully connected network, etc. [ABSTRACT FROM AUTHOR]

District heating networks (DHNs) are promising technologies to increase the efficiency and reduce emissions of heat distribution to residential and commercial buildings. The advent of the smart grid paradigm has introduced the usage of heating load forecasting tools in DHNs. They provide estimates of future heating load, improving the planning of heat production and power station maintenance. In this article, we propose a methodology based on the integrated use of regularized regression and clustering for generating predictive models of future heating load in DHNs. The methodology is tested on a real case study based on a dataset provided by AGSM, an Italian utility company that manages a DHN in the city of Verona, Italy. We generate a set of multiple-equation models having different degrees of complexity and show that models generated by the proposed approach outperform those trained by standard methods. Moreover, we provide an interpretation of patterns encoded by these models, and show that they identify real operational states of the network. The approach is completely data driven. [ABSTRACT FROM AUTHOR]

This article proposes an optimal energy scheduling method for power transmission networks using novel genetic algorithm (nGA) for solving the dynamic economic dispatch (DED) problem combined with machine learning based short-term load forecasting (STLF). The STLF is implemented based on a multilayer artificial neural network (MANN) to estimate the day-ahead variations in the demand. The efficacy of the proposed energy scheduling model together with the STLF is verified using a modified IEEE 30-bus system using real data of the power plants located in the Ereymentau region of Kazakhstan. The simulation results suggest that the proposed model offers a cost effective, reliable, and efficient dynamic energy scheduling in power transmission systems. [ABSTRACT FROM AUTHOR]

Virtual power plants (VPPs) have become a driving force for the decentralized energy industry, due to their efficient management and control of distributed energy resources. Most of the operation strategies for VPPs are designed based on the day-ahead forecasts. However, the prediction errors of the renewable energy sources (RES) and loads in the power dispatch schedule can lead to a suboptimal operation. In this article, an adaptive and predictive energy management strategy for a real-time optimal operation of VPPs is proposed based on the model predictive control technique with a feedback correction (FC) to compensate for the prediction error. This strategy has two parts: 1) receding-horizon optimization (RHO), and 2) FC. In the first part, a hybrid prediction algorithm based on the integration of the time-series model and the Kalman filter is used to forecast the output powers of RES and the loads. Based on the prediction, the RHO model schedules the operation following the latest forecast information. In the second part, the receding schedule is adjusted based on the fast-rolling gray model's ultrashort-term error prediction. The FC is applied to minimize the adjustments for compensating the prediction error. The proposed strategy is implemented on a VPP in a real electricity distribution system in New South Wales, Australia. The simulation results demonstrate the effectiveness of the proposed strategy with a better tracking of the actual available resources and a minimal mismatch between demand and supply. [ABSTRACT FROM AUTHOR]

Due to the visibility and controllability features of the latest distribution networks, there are potentials to model customer loads through a component-based approach more accurately. The electricity markets started in recent years enable trading and exchanges of electricity among traditional power stations and prosumers. It provides a possibility to improve the optimality of classical volt-var optimization (VVO) schemes. In this paper, a tri-level bottom-up load model and a market-based volt-var optimization (mVVO) model are proposed. The energy-to-voltage (ETV) parameter is proposed in our load model as a substitute for the widely used load-to-voltage (LTV) parameter to reflect long-term load characteristics. The mVVO model using ETV parameters aims to combine the optimal voltage regulation strategies and electricity purchase strategies from wholesale and retail markets for maximum cost saving. The case studies based on real load data of Hong Kong show that by coordinating the voltage regulation methods with market behaviors, utilities can reach a higher level of cost saving, in addition to energy saving. [ABSTRACT FROM AUTHOR]

Accurate and rapid load forecasting for industrial customers has been playing a crucial role in modern power systems. Due to the variability of industrial customers’ activities, individual industrial loads are usually too volatile to forecast accurately. In this paper, a short-term load forecasting model for industrial customers based on the Temporal Convolutional Network (TCN) and Light Gradient Boosting Machine (LightGBM) is proposed. Firstly, a fixed-length sliding time window method is adopted to reconstruct the electrical features. Next, the TCN is utilized to extract the hidden information and long-term temporal relationships in the input features including electrical features, a meteorological feature and date features. Further, a state-of-the-art LightGBM capable of forecasting industrial customers’ loads is adopted. The effectiveness of the proposed model is demonstrated by using datasets from different industries in China, Australia and Ireland. Multiple experiments and comparisons with existing models show that the proposed model provides accurate load forecasting results. [ABSTRACT FROM AUTHOR]

The predispatch price forecast plays a key element in the electricity market. However, such a forecast usually depends on the traditional offline batch-learning technologies, which cannot respond in time to the unexpected changes in the local power system environment. Further, the predispatch local price forecast is often affected by the dynamic price changes from the neighboring regions. This article proposes a novel online learning forecast approach to overcome the above issues to provide a better predispatch price forecast by using the online sequential extreme learning machine (OS-ELM) algorithm. The article proposes a novel data structure in the form of a 2-D orthogonal list and two corresponding OS-ELM modules. One module provides the rolling day-ahead price prediction and prediction intervals using the day-by-day online training update, while the other provides the rolling 30-min prediction using the 2-h-by-2-h online training update. The proposed approach can continuously perceive any unexpected events and any price fluctuations from the neighboring regions in the nonlinear patterns. The proposed approach is validated using simulation studies based on the data from the Australian electricity market, and the simulation results show that the proposed approach can help in improving the forecast accuracy, especially when unexpected changes occur both locally and in the neighboring area. [ABSTRACT FROM AUTHOR]

In this paper, previous theory and results regarding Low-Voltage (LV) distribution losses are validated through power-flow simulation using real load data with standard metering resolution. The validated results are then used to build a generalized loss prediction model for arbitrary grid topologies and metering resolutions. The paper provides the theoretical background to build and generalize a loss prediction model based on load variances together with an approach to estimate such variances. The prediction model uses as explanatory variables the variances of the loads. Yet, such variances are usually unknown for the higher metering resolutions, so the model also proposes a method to predict the higher resolution variances based on the standard metering ones. Results are presented to illustrate the approach proposed and to estimate the increase in losses expected for typical LV grids. [ABSTRACT FROM AUTHOR]

With the increase of solar photovoltaic (PV) permeability in power systems, high-precision PV power forecasting is critical to secure the economic operation of the smart grid. Solar irradiance is the most important factor affecting PV power. Accurately estimating the amount of solar irradiance leads to better production estimates from solar panels. Meanwhile, satellite images provide information on current and future cloud coverage, and are potentially useful in inferring solar irradiance. To accurately capture the effect of cloud on irradiance, this article develops a real-time mapping model between satellite image and solar irradiance. Based on the mapping model, a new ultrashort-term global horizontal irradiance (GHI) forecasting method combining satellite visible images and meteorological information is proposed. First, a modified method is utilized for reducing the influence of the solar zenith angle on pixel intensity and extracting the features in the image processing stage. Then, the cloud cover factors are extracted from satellite visible images by using the modified convolutional neural network. After that, the GHI forecasting model is developed, which is based on the combined use of meteorological information and cloud cover factors. Meanwhile, a cloud motion forecasting method using predicted wind speeds is developed for discussion. The better performance of the proposed method is demonstrated by comparing it with several benchmark models. [ABSTRACT FROM AUTHOR]

The smart meters installed at the customer premises are one of the main apparatuses promoting the modernization of the distribution systems. These devices collect a huge amount of data, demanding the development of analytic techniques to transform these data into useful information. In addition, by proposing new applications to data supplied by smart meters, more value is added to this equipment, allowing higher return on the associated investments. Utilities can have a business case by increasing their operational efficiency if smart meters are used, for instance, to identify non-technical losses, which represent an important cause of revenue losses. This paper presents a new data analytic technique for detection and location of non-technical losses caused by illegal connections of loads to distribution systems in the presence of smart meters. The data analytic technique relies on bad data analysis, similar to the ones used in state estimation methods, developed specifically for this application. A real 34-bus low voltage system is used to illustrate the main concepts of the proposed algorithm. Systematic tests are also conducted on a real 1682-bus distribution system to evaluate the method performance considering electricity theft caused by medium and low voltage customers. [ABSTRACT FROM AUTHOR]

The unscented Kalman filter (UKF) provides a powerful tool for power system forecasting-aided state estimation (FASE). However, when the power systems are affected by the abnormal operating situations, i.e., the non-Gaussian communication noises, sudden loads or state changes, and instrument failures, the original UKF based on the minimum mean square error (MMSE) criterion may suffer from performance degradation. In contrast to the MMSE criterion, the minimum error entropy (MEE) exhibits the robustness with respect to complex non-Gaussian disturbances. In this article, we develop a new unscented Kalman-type filter based on the MEE criterion, termed MEE-UKF. To derive the MEE-UKF, a statistical linearization approach is adopted in the augmented model such that the state and measurement errors are combined in the MEE cost function simultaneously. Then, a fixed-point iteration algorithm is used to recursively update the posterior estimates and covariance matrix. Apart from the impulsive noises, the MEE-UKF can deal with complex multimodal distribution noises in both process and measurement. The high accuracy and strong robustness of MEE-UKF are confirmed by the simulation results on IEEE 14, 30, and 57 bus test systems under different non-Gaussian disturbances. [ABSTRACT FROM AUTHOR]

In this work we address the problem of short-term load forecasting. We propose a generalization of the linear state-space model where the evolution of the state and the observation matrices is unknown. The proposed blind Kalman filter algorithm proceeds via alternating the estimation of these unknown matrices and the inference of the state, within the framework of expectation-maximization. A mini-batch processing strategy is introduced to allow on-the-fly forecasting. The experimental results show that the proposed method outperforms the state-of-the-art techniques by a considerable margin, both on load profile estimation and peak load forecast problems. [ABSTRACT FROM AUTHOR]

A new multiple parallel input and parallel output architecture-based models are developed for forecasting electric load power consumption. Attention was paid toward the improvement of the accuracy of forecasting using this new architecture built-in by us, as compared to others in the latest literature. The models have an ability to forecast daily load profiles with a lead time of one to seven days. Both multivariate linear regression and feed-forward neural network techniques have been chosen for comparative performance study and analysis. The real-time data used for this research work were collected from Tata Power Delhi Distribution Limited, Delhi, India. Based on the performance criteria provided in the literature, each model is analyzed and the results are presented for two different lead times, i.e., day-ahead and week-ahead only. [ABSTRACT FROM AUTHOR]

Phase-locked loops (PLLs) are widely popular in applications of grid synchronization for single- and three-phase utility power systems. However, these PLL-based methods are usually designed for constant frequency grids and these schemes only have fast dynamics and good accuracy in a small local region (i.e., phase-locked region). When applying these PLLs in the grids of more electric aircraft (MEA), in which the grid frequency varies in a wide range (360–800 Hz), both the dynamic response and the steady-state accuracy of PLLs reduce significantly. In light of this, this article presents a global convergence method to estimate the frequency, positive- and negative-sequence components of grid voltage in MEA power systems. By analyzing the relationship between the DC-offset component, the grid voltage, and their derivatives, a mathematic model with the unknown parameters, which are the DC-offset component and the grid frequency, is established. Accordingly, a global estimator is proposed to estimate the grid voltage parameters accurately. A Lyaponov-based argument ensures that the proposed estimator is global convergent and it has the capability to eliminate the DC offset effect completely. The experimental results are provided to verify the effectiveness of the proposed estimator. [ABSTRACT FROM AUTHOR]

Inspired from recent insights into the common ground of machine learning, optimization and decision-making, this paper proposes an easy-to-implement, but effective procedure to enhance both the quality of renewable energy forecasts and the competitive edge of renewable energy producers in electricity markets with a dual-price settlement of imbalances. The quality and economic gains brought by the proposed procedure essentially stem from the utilization of valuable predictors (also known as features) in a data-driven newsvendor model that renders a computationally inexpensive linear program. We illustrate the proposed procedure and numerically assess its benefits on a realistic case study that considers the aggregate wind power production in the Danish DK1 bidding zone as the variable to be predicted and traded. Within this context, our procedure leverages, among others, spatial information in the form of wind power forecasts issued by transmission system operators (TSO) in surrounding bidding zones and publicly available in online platforms. We show that our method is able to improve the quality of the wind power forecast issued by the Danish TSO by several percentage points (when measured in terms of the mean absolute or the root mean square error) and to significantly reduce the balancing costs incurred by the wind power producer. [ABSTRACT FROM AUTHOR]

The workflows of the predominant datacenter services are underlaid by various Fork-Join structures. Due to the lack of good understanding of the performance of Fork-Join structures in general, today's datacenters often operate under low resource utilization to meet stringent service level objectives (SLOs), e.g., in terms of tail and/or mean latency, for such services. Hence, to achieve high resource utilization, while meeting stringent SLOs, it is of paramount importance to be able to accurately predict the tail and/or mean latency for a broad range of Fork-Join structures of practical interests. In this article, we propose a black-box Fork-Join model that covers a wide range of Fork-Join structures for the prediction of tail and mean latency, called ForkTail and ForkMean, respectively. We derive highly computational effective, empirical expressions for tail and mean latency as functions of means and variances of task response times. Our extensive testing results based on model-based and trace-driven simulations, as well as a real-world case study in a cloud environment demonstrate that the models can consistently predict the tail and mean latency within 20 and 15 percent prediction errors at 80 and 90 percent load levels, respectively, for heavy-tailed workloads, and at any load levels for light-tailed workloads. Moreover, our sensitivity analysis demonstrates that such errors can be well compensated for with no more than 7 percent resource overprovisioning. Consequently, the proposed prediction model can be used as a powerful tool to aid the design of tail-and-mean-latency guaranteed job scheduling and resource provisioning, especially at high load, for datacenter applications. [ABSTRACT FROM AUTHOR]

The increasing deployment of triggerless power quality monitors provides valuable data about power systems and their components. However, efficient and accurate data mining techniques are necessary for parsing through large datasets and obtaining useful insights about the power system operation. While multiple studies can be found in the literature for detection of power quality events, they either detect only severe disturbances or are applicable to a specific subset of disturbances. This study proposes a universal approach for detecting power quality disturbances within large datasets regardless of the event root causes. It is based on the comparison between the shapes of multiple cycles of waveform data through similarity scores. Pre-processing techniques to enhance the detector accuracy are presented. They include correction of time misalignment between successive cycles caused by variations in the power system operating frequency, and reduction of load variation effects. The threshold values for detecting abnormal waveforms is time-adaptive, so that detection is based on local outliers. The proposed framework performance is assessed through multiple types of field datasets, and its superiority over other waveform abnormality detectors is empirically demonstrated. Moreover, current waveform is shown to be more suitable than that of voltage for novelty detection in power quality data. [ABSTRACT FROM AUTHOR]

To maintain stable and reliable operations in smart grid, accurate state estimation is of paramount importance. However, synthesized false data injection attacks could wisely circumvent conventional bad data detection mechanisms by introducing arbitrary errors to state estimates to seriously affect the entire power system operation. To defend these attacks, phase measurement units (PMUs) are deployed in advance at various locations to reduce the chance of being attacked. However, when the budget of placement is not large enough so that the whole system cannot be covered by PMUs, the existing PMU placement algorithms based on greedy strategies are insufficient in some weak locations due to the nature of greedy strategies. In this article, we propose a new hybrid attack, which can be easily used by attackers to attack the buses with less connectivity and impose adverse impacts to state estimation with a low-attack cost so that existing defenses based on greedy strategies become invalid. We future propose a predeployment PMU greedy algorithm for this new attack in which the most vulnerable buses are first protected and, then, a greedy-based algorithm is used to deploy other PMUs until the whole system is observable. Experimental results on various IEEE standard systems demonstrate the effectiveness of our schemes. [ABSTRACT FROM AUTHOR]

Short-term household electricity load forecasting is important for utility companies to ensure reliable power supplies. Traditional methods for load forecasting relied on historical records from one single data source and have limitations with insufficient or missing data. Recently, an emerging family of machine learning algorithms, multitask learning (MTL), has been developed and has the potential for load forecasting. By MTL, the electricity consumption data from multiple communities can be fused to improve forecasting accuracy. However, appropriate modeling of the relatedness to enable the between-community knowledge transfer remains a challenge. This paper proposes an improved MTL algorithm for a Bayesian spatiotemporal Gaussian process model (BSGP) to characterize the relatedness among the different residential communities. It hypothesizes on the similar impacts of environmental and traffic conditions on electricity consumption in improving short-term load forecasting. Furthermore, this paper proposes a low-ranked dirty model along with an iterative algorithm to improve the learning of model parameters under an MTL framework. This paper uses a real-world case study from two residential communities in Tallahassee, Fl, USA, to demonstrate the method effectiveness. The proposed method significantly outperforms state-of-the-art forecasting methods and effectively captures the relatedness to provide between-community knowledge transfer compared with other MTL methods. [ABSTRACT FROM AUTHOR]

This article presents a new approach for estimating the frequency variations ($Df$) due to applying smart grid functions on a load bus. The proposed approach is based on constructing a ZIP model for the power demands at a load bus that is controlled by smart grid functions.The constructed ZIP model provides a relationship between the active and reactive power demands and the frequency at any load bus. This relationship can be formulated as a set of nonlinear equations, which can be numerically solved for $Df$. The load-model approach is implemented for performance evaluation using load buses in the IEEE 30-bus power system under different loading levels. Performance results show that the proposed approach has a simple implementation, and can provide an accurate estimation of frequency variations due to slow and small changes load power demands. Furthermore, performance results reveal the insensitivity of the load-model approach to load power demands and/or seasonal changes in load power demands. [ABSTRACT FROM AUTHOR]