Long term load forecasting with hourly predictions based on long-short-term-memory networks

Agrawal, Rahul; Muchahary, Frankle; Tripathi, M. M.

doi:10.1109/tpec.2018.8312088

Cited by 93 publications

(45 citation statements)

References 12 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In terms of aggregated (substation) level, the aggregation of all individual forecasts yielded better results than the direct forecast of aggregated loads. Agrawal et al [32] Introduced a deep-structure RNN-LSTM network at a higher aggregation level; ISO New England energy market using daily, monthly and weekly features to produce hourly predictions over a one-year period. Similar to the approaches using shallow ANNs, some studies explored the combination of LSTM with other models or optimization algorithms.…”

Section: Related Workmentioning

confidence: 99%

Short-Term Load Forecasting Using Smart Meter Data: A Generalization Analysis

et al. 2020

View full text Add to dashboard Cite

Short-term load forecasting ensures the efficient operation of power systems besides affording continuous power supply for energy consumers. Smart meters that are capable of providing detailed information on buildings energy consumption, open several doors of opportunity to short-term load forecasting at the individual building level. In the current paper, four machine learning methods have been employed to forecast the daily peak and hourly energy consumption of domestic buildings. The utilized models depend merely on buildings historical energy consumption and are evaluated on the profiles that were not previously trained on. It is evident that developing data-driven models lacking external information such as weather and building data are of great importance under the situations that the access to such information is limited or the computational procedures are costly. Moreover, the performance evaluation of the models on separated house profiles determines their generalization ability for unseen consumption profiles. The conducted experiments on the smart meter data of several UK houses demonstrated that if the models are fed with sufficient historical data, they can be generalized to a satisfactory level and produce quite accurate results even if they only use past consumption values as the predictor variables. Furthermore, among the four applied models, the ones based on deep learning and ensemble techniques, display better performance in predicting daily peak load consumption than those of others.

show abstract

Section: Related Workmentioning

confidence: 99%

Short-Term Load Forecasting Using Smart Meter Data: A Generalization Analysis

et al. 2020

View full text Add to dashboard Cite

show abstract

“…It indicates that forecasting models based on LSTM outperform other methods, including auto-regressive integrated moving average (ARIMA), support vector regression (SVR), and traditional feed-forward neural network (FNN) [20]. Kumar et al used the LSTM model for short-term electric load forecasting using twelve years of historical data for ISO New England electricity market and reported very reliable and robust results [21]. Zheng et al proposed a short-term load forecasting method for residential community based on gated recurrent unit neural network.…”

Section: Literature Reviewmentioning

confidence: 99%

Short-Term Load Forecasting Using Recurrent Neural Networks With Input Attention Mechanism and Hidden Connection Mechanism

Zhang

2020

IEEE Access

View full text Add to dashboard Cite

Short-term load forecasting is a critical task in the smart grid, which can be used to optimize power deployment and reduce power losses. Recurrent neural networks (RNNs) are the most popular deep learning models for short-term load forecasting. However, despite of achieving better forecasting accuracy than the traditional models, the performance of the existing RNN-based load forecasting approaches is still unsatisfactory for practical usage. Therefore, in this work, we have proposed input attention mechanism (IAM) and hidden connection mechanism (HCM) to greatly enhance the accuracy and efficiency of RNNbased load forecasting models. Specifically, we use IAM to assign the importance weights on input layers, which have better performances in both efficiency and accuracy than traditional attention mechanisms. To further enhance the models' efficiency, HCM is then applied to utilize residual connections to enhance the model's converging speed. We have applied both IAM and HCM on four state-of-the-art RNN implementations, and then conducted extensive experimental studies on two public datasets. Experimental results show that the proposed RNNs with IAM and HCM models achieve much better performances than the state-of-the-art baselines in both accuracy and efficiency. Ablation studies show that both IAM and HCM are essential to achieve such superior performances.

show abstract

“…Similarly, if there exists more than one home user, then even predicting short-term load in the form of regular and irregular child nodes could result in poor performance. In this regard, the long-term load prediction and forecasting models are employed to overcome the challenges present in the short-term forecasting models [ 25 , 26 ]. In many of these research works; the LSTM model is widely adopted for forecasting long-term load.…”

Section: Related Workmentioning

confidence: 99%

Towards Energy Efficient Home Automation: A Deep Learning Approach

Khan

Seo

Kim

2020

Sensors

View full text Add to dashboard Cite

Home Automation Systems (HAS) attracted much attention during the last decade due to the developments in new wireless technologies, such as Bluetooth 4.0, 5G, WiFi 6, etc. In order to enable automation as a service in smart homes, a number of challenges must be addressed, such as fulfilling the electrical energy demands, scheduling the operational time of appliances, applying machine learning models in real-time, optimal human appliances interaction, etc. In order to address the aforementioned challenges and control the wastage of energy due to the lifestyle of the home users, we propose a system for automatically controlling the energy consumption by employing machine and deep learning techniques to smart home networks. The proposed system works in three phases, (1) feature extraction and classification based on 1-dimensional Deep Convolutional Neural Network (1D-DCNN) which extract important energy patterns from the historic energy data, (2) a load forecasting system based on Long-short Term Memory (LSTM) is proposed to forecast the load based on the extracted features in phase 1 and (3) a scheduling algorithm based on the forecasted data obtained from phase 2 is designed to schedule the operational time of smart home appliances. The proposed scheme efficiently automates the smart home appliances to consume less energy while adapting to the lifestyle of smart home users. The validation of the proposed scheme is tested with a number of simulation scenarios incorporating datasets from authentic data sources. The simulation results show that the proposed smart home automation system can be a game-changer in fulfilling the energy demands of the home users without installing renewable and other energy sources in the future.

show abstract

Long term load forecasting with hourly predictions based on long-short-term-memory networks

Cited by 93 publications

References 12 publications

Short-Term Load Forecasting Using Smart Meter Data: A Generalization Analysis

Short-Term Load Forecasting Using Smart Meter Data: A Generalization Analysis

Short-Term Load Forecasting Using Recurrent Neural Networks With Input Attention Mechanism and Hidden Connection Mechanism

Towards Energy Efficient Home Automation: A Deep Learning Approach

Contact Info

Product

Resources

About