Short-term Residential Load Research Articles

MLFGCN: short-term residential load forecasting via graph attention temporal convolution network.

Residential load forecasting is a challenging task due to the random fluctuations caused by complex correlations and individual differences. The existing short-term load forecasting models usually introduce external influencing factors such as climate and date. However, these additional information not only bring computational burden to the model, but also have uncertainty. To address these issues, we propose a novel multi-level feature fusion model based on graph attention temporal convolutional network (MLFGCN) for short-term residential load forecasting. The proposed MLFGCN model fully considers the potential long-term dependencies in a single load series and the correlations between multiple load series, and does not require any additional information to be added. Temporal convolutional network (TCN) with gating mechanism is introduced to learn potential long-term dependencies in the original load series. In addition, we design two graph attentive convolutional modules to capture potential multi-level dependencies in load data. Finally, the outputs of each module are fused through an information fusion layer to obtain the highly accurate forecasting results. We conduct validation experiments on two real-world datasets. The results show that the proposed MLFGCN model achieves 0.25, 7.58% and 0.50 for MAE, MAPE and RMSE, respectively. These values are significantly better than those of baseline models. The MLFGCN algorithm proposed in this paper can significantly improve the accuracy of short-term residential load forecasting. This is achieved through high-quality feature reconstruction, comprehensive information graph construction and spatiotemporal features capture.

Short-Term Residential Load Forecasting via Pooling-Ensemble Model With Smoothing Clustering

Short-Term Residential Load Forecasting via Pooling-Ensemble Model With Smoothing Clustering

Enhancing multivariate, multi-step residential load forecasting with spatiotemporal graph attention-enabled transformer

Short-term residential load forecasting (STRLF) holds great significance for the stable and economic operation of distributed power systems. Different households in the same region may exhibit similar consumption patterns owing to the analogous environmental parameters. Incorporating the spatiotemporal correlations can enhance the load forecasting performance of individual households. To this end, a spatiotemporal graph attention (STGA)-enabled Transformer is proposed for multivariate, multi-step residential load forecasting in this paper. Specifically, the multiple residential loads are cast to a graph and a Transformer with a graph sequence-to-sequence (Seq2Seq) structure is employed to model the multi-step load forecasting problem. Gated fusion-based STGA blocks are embedded in the encoder and decoder of the Transformer to extract dynamic spatial correlations and non-linear temporal patterns among multiple residential loads. A transform attention block is further designed to transfer historical graph observations into future graph predictions and alleviate the error propagation between the encoder and decoder. The embedding of multiple attention modules in the Seq2Seq framework allows us to capture the spatiotemporal correlations between residents and achieve confident inference of load values several steps ahead. Numerical simulations on residential data from three different regions demonstrate that the developed Transformer method improves multi-step load forecasting by 14.7% at least, compared to the state-of-the-art benchmarks.

GRU combined model based on multi-objective optimization for short-term residential load forecasting

Reliable and accurate short-term forecasting of residential load plays an important role in DSM. However, the high uncertainty inherent in single-user loads makes them difficult to forecast accurately. Various traditional methods have been used to address the problem of residential load forecasting. A single load forecast model in the traditional method does not allow for comprehensive learning of data characteristics for residential loads, and utilizing RNNs faces the problem of long-term memory with vanishing or exploding gradients in backpropagation. Therefore, a gated GRU combined model based on multi-objective optimization is proposed to improve the short-term residential load forecasting accuracy in this paper. In order to demonstrate the effectiveness, GRUCC-MOP is first experimentally tested with the unimproved model to verify the model performance and forecasting effectiveness. Secondly the method is evaluated experimentally with other excellent forecasting methods: models such as DBN, LSTM, GRU, EMD-DBN and EMD-MODBN. By comparing simulation experiments, the proposed GRU combined model can get better results in terms of MAPE on January, April, July, and November load data, so this proposed method has better performance than other research methods in short-term residential load forecasting.

Enhancing peak prediction in residential load forecasting with soft dynamic time wrapping loss functions

Short-term residential load forecasting plays a crucial role in smart grids, ensuring an optimal match between energy demands and generation. With the inherent volatility of residential load patterns, deep learning has gained attention due to its ability to capture complex nonlinear relationships within hidden layers. However, most existing studies have relied on default loss functions such as mean squared error (MSE) or mean absolute error (MAE) for neural networks. These loss functions, while effective in overall prediction accuracy, lack specialized focus on accurately predicting load peaks. This article presents a comparative analysis of soft-DTW loss function, a smoothed formulation of Dynamic Time Wrapping (DTW), compared to other commonly used loss functions, in order to assess its effectiveness in improving peak prediction accuracy. To evaluate peak performance, we introduce a novel evaluation methodology using confusion matrix and propose new errors for peak position and peak load, tailored specifically for assessing peak performance in short-term load forecasting. Our results demonstrate the superiority of soft-DTW in capturing and predicting load peaks, surpassing other commonly used loss functions. Furthermore, the combination of soft-DTW with other loss functions, such as soft-DTW + MSE, soft-DTW + MAE, and soft-DTW + TDI (Time Distortion Index), also enhances peak prediction. However, the differences between these combined soft-DTW loss functions are not substantial. These findings highlight the significance of utilizing specialized loss functions, like soft-DTW, to improve peak prediction accuracy in short-term load forecasting.

Short-Term Residential Load Forecasting With Baseline-Refinement Profiles and Bi-Attention Mechanism

Short-Term Residential Load Forecasting With Baseline-Refinement Profiles and Bi-Attention Mechanism

Privacy-Preserving and Hierarchically Federated Framework for Short-Term Residential Load Forecasting

Short-term Load Forecasting (STLF) plays a fundamental role in modern energy systems. This paper proposes a privacy-preserving STLF framework for residential energy users. The framework enables users to collaboratively train an STLF model without exchanging their load data. The system works in two stages: (i) a federated user clustering method is developed to divide the users into multiple clusters based on load pattern similarity. Instead of aggregating the users’ data to do clustering, the developed method determines the user clusters in a distributed manner. (ii) after the user clustering stage, a hierarchically federated STLF model training is developed, which applies federated learning principles to facilitate intra-and inter-user cluster model training. In the process, the individual users also do not expose their load data. Further, an asynchronous communication mechanism is designed and integrated into the framework, making it high fault-tolerant and adaptable to the residential environment with communication uncertainties. Comprehensive experiments based on a real Australian load dataset are conducted to validate the system. The simulation results show that the proposed framework can effectively train the load forecasting model with a fast coverage rate and achieve up to 37.25% prediction accuracy improvement (in terms of the Mean Absolute Percentage Error metric) compared with the other benchmark methods.

Interpretable transformer-based model for probabilistic short-term forecasting of residential net load

Short-term residential load forecasting is of great significance for the demand-side energy management of the grid and the home energy management of resident customers. The massive penetration of distributed renewable energy, especially small-scale solar photovoltaic (PV), in the residential sector urgently requires us to move from traditional load forecasting to net load forecasting. In recent years, deep learning techniques that can improve model forecasting performance have developed rapidly in the field of residential load forecasting. However, the opaque nature of deep learning makes its practical application very difficult. This paper proposes a Transformer-based probabilistic residential net load forecasting method that utilizes quantile regression to quantify uncertainty in future load demand. Meanwhile, to improve the interpretability of deep learning model, local variable selection network is developed to automatically select relevant features and provide feature-level explanations. Additionally, interpretable sparse self-attention mechanism is proposed to extract long-term temporal dependencies. Numerical experiments are carried out with data from real household smart meters. The results show that the proposed model outperforms other state-of-the-art forecasting models. In terms of point forecasting, compared with the most common deep time series forecasting model LSTM, the proposed model decreases by 21.3%, 27.3% and 20.9% On three point forecasting performance metrics. Compared with Vanilla Transformer, the proposed InterFormer decreases by 9.9%, 10.5% and 9.0% On three point forecasting performance metrics. In terms of probabilistic forecasting, compared with LSTM and Vanilla Transformer, the average pinball loss of the proposed model decreases by 26.2% and 17.0%, respectively. Additionally, and most importantly, our model provides users with explanations in terms of feature importance and temporal patterns.

A residential load forecasting method for multi-attribute adversarial learning considering multi-source uncertainties

The rapid development of the Internet of Things and device-level meters provides accurate energy consumption data for various household devices, and making full use of these data to achieve higher accuracy of residential load forecasting is an open problem. Moreover, the uncertainties and insufficient feature extraction of variables make it challenging for individual residential load forecasts. In this regard, this paper proposes a novel short-term residential load forecasting method. It has the following aspects: (1) Considering the multi-source uncertainties, a multi-source uncertainties divide-and-conquer mechanism is proposed, which classifies the residential load from the device level according to different sources of uncertainties and forecasts separately. (2) Considering the multiple attributes of influencing variables, a multi-attribute adversarial learning mechanism based on the conditional Wasserstein generative adversarial network with gradient penalty (cWGAN-GP) is designed. In cWGAN-GP, this paper uses a long short-term memory network to implement the generator and uses a convolutional neural network to implement the discriminator. The former is used to extract temporal features, while the latter is used to extract spatial features, and they realize the fusion of multiple attributes through adversarial training to improve prediction accuracy. Compared with existing load forecasting methods, our proposed method’s mean absolute percentage error decreased by 1.5–38.9%, mean square error decreased by 3.5–35.9%, normalized root MSE decreased by 1.7–19.9%, and mean absolute error decreased by 3.7–27.0%.

A short-term residential load forecasting scheme based on the multiple correlation-temporal graph neural networks

Accurate residential load forecasting (RLF) is of great significance for the decision-making and operation of modern power system. In literature, deep neural network (DNN) based RLF schemes have witnessed great development due to the advantage of automatically extracting features and capturing complex non-linear pattern in the presented data. However, most existed works separately exploit the historical data of a specific residential house to forecast its load. However, the electricity consumption behaviors among residential users are not independent, and implicitly have some correlations, which can be explicitly characterized and exploited to improve the accuracy of RLF forecasting. Inspired by this idea, through exploiting the multiple correlations among households, this paper proposes a novel residential load forecasting framework based on multiple correlation-temporal graph neural networks, RLF-MGNN. Specifically, the novelty of our work includes three aspects. First, multiple graphs are explicitly constructed to represent both linear and nonlinear correlations among temporal load series of households. That is, the synchronization graph is built to describe the degree of linear correlation between two households using Pearson correlation coefficient, which characterizes the similarity of their consumption behaviors, and the causality graph is built to describe their nonlinear correlation using transfer entropy, which characterizes the amount of directional information transfer from one time series to another, and models the mutual influence between households. Second, the multiple correlation-temporal graph convolutional networks (GCNs) are designed to forecast the residential users’ loads. In detail, at each timestep, latent features are first extracted by corresponding GCNs to embed multiple correlations among households, and then are sent to Long Short-Term Memory (LSTM) for further learning the latent temporal features. Finally, thorough experiments on real datasets demonstrate that our proposed RLF-MGNN outperforms the state-of-the-art independent DNN based schemes and other GNN based schemes.

Spatial and Temporal Attention-Enabled Transformer Network for Multivariate Short-Term Residential Load Forecasting

Spatial and Temporal Attention-Enabled Transformer Network for Multivariate Short-Term Residential Load Forecasting

Federated learning for interpretable short-term residential load forecasting in edge computing network

Federated learning for interpretable short-term residential load forecasting in edge computing network

Very short-term residential load forecasting based on deep-autoformer

Very short-term load forecasting (VSLTF) plays an essential role in guaranteeing effective electricity dispatching and generating in residential microgrid systems. However, the extreme fluctuations and irregular data patterns of VSTLF have brought severe challenges to accurate forecasting. Deep learning methods have been mostly utilized in time series predicting tasks like load forecasting. Recently, an Autoformer neural network has been proposed in many time series forecasting scenarios. Based on Autoformer, this paper proposes a new Deep-Autoformer framework, where the extra MLP layers are added to the basic Autoformer framework for a more efficient deep information extraction. Taking a microgrid system in Austin, Texas from the Pecan Street dataset as a case study, Deep-Autoformer and other five baseline models are utilized to forecast the load data of 15 min and one-hour time resolution. The main contributions of the proposed Deep-Autoformer are: (i)the experiment results indicate that the proposed Deep-Autoformer has achieved State-Of-The-Art (SOTA) results in both VSTLF and STLF, and(ii) the ‘deep’ method, where the MLP layers are added in the appropriate positions of the neural network, can contribute to more efficient feature extraction. Moreover, due to the unintuitive phenomenons in the experiment, two hypotheses are also proposed: (i) the long-ago historical data may affect the performance of the auto-correlation mechanism of the Autoformer, and (ii) models are probably overfitting the historical patterns if the time series data are too long. Overall, the proposed Deep-Autoformer can provide a feasible approach and a new baseline for the real application of VSTLF.

Short term residential load forecasting using long short-term memory recurrent neural network

<span>Load forecasting plays an essential role in power system planning. The efficiency and reliability of the whole power system can be increased with proper planning and organization. Residential load forecasting is indispensable due to its increasing role in the smart grid environment. Nowadays, smart meters can be deployed at the residential level for collecting historical data consumption of residents. Although the employment of smart meters ensures large data availability, the inconsistency of load data makes it challenging and taxing to forecast accurately. Therefore, the traditional forecasting techniques may not suffice the purpose. However, a deep learning forecasting network-based long short-term memory (LSTM) is proposed in this paper. The powerful nonlinear mapping capabilities of RNN in time series make it effective along with the higher learning capabilities of long sequences of LSTM. The proposed method is tested and validated through available real-world data sets. A comparison of LSTM is then made with two traditionally available techniques, exponential smoothing and auto-regressive integrated moving average model (ARIMA). Real data from 12 houses over three months is used to evaluate and validate the performance of load forecasts performed using the three mentioned techniques. LSTM model has achieved the best results due to its higher capability of memorizing large data in time series-based predictions.</span>

Short-term residential load forecasting using Graph Convolutional Recurrent Neural Networks

The abundance of energy consumption data collected by smart meters has inspired researchers to employ deep neural networks to solve the existing problems in the power industry, such as Short-Term Load Forecasting (STLF). Most studies addressing the STLF problem, focus on historical load data and to achieve higher performance, they supplement costly accessible environmental and calendar variables with data. This approach ignores the existing spatial information among the consumers which subsequently might lead into the emergence of similar consumption patterns. In this paper, we present a Graph Convolutional Recurrent Neural Network, a novel neural architecture, for STLF problem that combines Graph Convolutional Networks and Long Short-Term Memory networks to simultaneously extract spatial and temporal information from users with similar consumption patterns. Our model captures spatial information from users without prior knowledge of their geographic location and does not rely on additional environmental variables. We compared our model to traditional baseline models for STLF using two real-world electricity consumption datasets. The empirical results demonstrate a significant improvement in prediction compared with the baseline models, exhibiting a 9.5% and an 8% improvement in terms of Mean Absolute Percentage Error, in the Customer Behavior Trials and Low Carbon London datasets, respectively.

Impacts of traffic data on short-term residential load forecasting before and during the COVID-19 pandemic

Accurate load forecasting is essential for power-sector planning and management. This applies during normal situations as well as phase changes such as the Coronavirus (COVID-19) pandemic due to variations in electricity consumption that made it difficult for system operators to forecast load accurately. So far, few studies have used traffic data to improve load prediction accuracy. This paper aims to investigate the influence of traffic data in combination with other commonly used features (historical load, weather, and time) – to better predict short-term residential electricity consumption. Based on data from two selected distribution grid areas in Switzerland and random forest as a forecasting technique, the findings suggest that the impact of traffic data on load forecasts is much smaller than the impact of time variables. However, traffic data could improve load forecasting where information on historical load is not available. Another benefit of using traffic data is that it might explain the phenomenon of interest better than historical electricity demand. Some of our findings vary greatly between the two datasets, indicating the importance of studies based on larger numbers of datasets, features, and forecasting approaches.

Federated Learning for Short-Term Residential Load Forecasting

Load forecasting is an essential task performed within the energy industry to help balance supply with demand and maintain a stable load on the electricity grid. As supply transitions towards less reliable renewable energy generation, smart meters will prove a vital component to facilitate these forecasting tasks. However, smart meter adoption is low among privacy-conscious consumers that fear intrusion upon their fine-grained consumption data. In this work we propose and explore a federated learning (FL) based approach for training forecasting models in a distributed, collaborative manner whilst retaining the privacy of the underlying data. We compare two approaches: FL, and a clustered variant, FL+HC against a non-private, centralised learning approach and a fully private, localised learning approach. Within these approaches, we measure model performance using RMSE and computational efficiency. In addition, we suggest the FL strategies are followed by a personalisation step and show that model performance can be improved by doing so. We show that FL+HC followed by personalisation can achieve a ~5% improvement in model performance with a ~10x reduction in computation compared to localised learning. Finally we provide advice on private aggregation of predictions for building a private end-to-end load forecasting application.

A Novel Short-Term Residential Electric Load Forecasting Method Based on Adaptive Load Aggregation and Deep Learning Algorithms

Short-term residential load forecasting is the precondition of the day-ahead and intra-day scheduling strategy of the household microgrid. Existing short-term electric load forecasting methods are mainly used to obtain regional power load for system-level power dispatch. Due to the high volatility, strong randomness, and weak regularity of the residential load of a single household, the mean absolute percentage error (MAPE) of the traditional methods forecasting results would be too big to be used for home energy management. With the increase in the total number of households, the aggregated load becomes more and more stable, and the cyclical pattern of the aggregated load becomes more and more distinct. In the meantime, the maximum daily load does not increase linearly with the increase in households in a small area. Therefore, in our proposed short-term residential load forecasting method, an optimal number of households would be selected adaptively, and the total aggregated residential load of the selected households is used for load prediction. In addition, ordering points to identify the clustering structure (OPTICS) algorithm are also selected to cluster households with similar power consumption patterns adaptively. It can be used to enhance the periodic regularity of the aggregated load in alternative. The aggregated residential load and encoded external factors are then used to predict the load in the next half an hour. The long short-term memory (LSTM) deep learning algorithm is used in the prediction because of its inherited ability to maintain historical data regularity in the forecasting process. The experimental data have verified the effectiveness and accuracy of our proposed method.

Spatial-Temporal Residential Short-Term Load Forecasting via Graph Neural Networks

Electric load forecasting, especially short-term load forecasting, is of significant importance for the safe and efficient operation of power grids. With the wide adoption of advanced smart meters, more attention has been paid to short-term residential load forecasting. Most of the existing load forecasting methods are mainly focused on using temporal information of historical loads, and information of neighboring houses are generally ignored. However, houses in the same or neighboring areas may show similar consumption patterns due to shared conditions such as temperature, holiday impacts. Such information can be very helpful for machine learning based forecasting methods. In this paper, we propose to tackle the short-term residential load forecasting including both the individual load and aggregated load with a graph neural network based forecasting framework. The proposed framework can capture the hidden spatial dependencies of different houses without even any prior knowledge requirement on the geographic information for these houses. The proposed framework is evaluated on data sets of different residential houses from several areas. The experimental results demonstrate that the proposed framework can improve the residential forecasting accuracy by a wide margin compared with the baselines.

Deep reservoir architecture for short-term residential load forecasting: An online learning scheme for edge computing

The short-term residential electricity demand forecast tasks are essential for designing the portfolio of the decision-making process of aggregators for the selection of control targets and derivation of specific control target values for demand response. The implementation of edge computing on the forecast functionality, which provides forecasting and minimal communication functions based on the data accumulated at the end-user-side using an inexpensive computational resource deployed locally, is an attractive framework for practical implementation. To achieve good accuracy while forecasting the demands of individual households during practical long-term operation, it is necessary to follow the changes in demand characteristics quickly by updating the prediction models frequently. This study focused on the concept of deep reservoir computing, which has a high affinity for implementation in edge computing frameworks from the viewpoint of hardware implementability, while utilizing the deep architecture and avoiding high computational cost. In particular, this study proposed an online learning scheme for the implementation of edge computing, which does not require large-scale computation for updating the prediction models. The effectiveness of the proposed approach was evaluated considering the calculation cost and accuracy through numerical experiments using real-world residential load data.