ﻻ يوجد ملخص باللغة العربية
Modern smart distribution system requires storage, transmission and processing of big data generated by sensors installed in electric meters. On one hand, this data is essentially required for intelligent decision making by smart grid but on the other hand storage, transmission and processing of that huge amount of data is also a challenge. Present approaches to compress this information have only relied on the traditional matrix decomposition techniques benefitting from low number of principal components to represent the entire data. This paper proposes a cascaded data compression technique that blends three different methods in order to achieve high compression rate for efficient storage and transmission. In the first and second stages, two lossy data compression techniques are used, namely Singular Value Decomposition (SVD) and Normalization; Third stage achieves further compression by using the technique of Sparsity Encoding (SE) which is a lossless compression technique but only having appreciable benefits for sparse data sets. Our simulation results show that the combined use of the 3 techniques achieves data compression ratio to be 15% higher than state of the art SVD for small, sparse datasets and up to 28% higher in large, non-sparse datasets with acceptable Mean Absolute Error (MAE).
The conventional approach to pre-process data for compression is to apply transforms such as the Fourier, the Karhunen-Lo`{e}ve, or wavelet transforms. One drawback from adopting such an approach is that it is independent of the use of the compressed
Detecting inaccurate smart meters and targeting them for replacement can save significant resources. For this purpose, a novel deep-learning method was developed based on long short-term memory (LSTM) and a modified convolutional neural network (CNN)
We introduce a conceptually simple and scalable framework for continual learning domains where tasks are learned sequentially. Our method is constant in the number of parameters and is designed to preserve performance on previously encountered tasks
Sequential recommender systems (SRS) have become the key technology in capturing users dynamic interests and generating high-quality recommendations. Current state-of-the-art sequential recommender models are typically based on a sandwich-structured
The recent advent of smart meters has led to large micro-level datasets. For the first time, the electricity consumption at individual sites is available on a near real-time basis. Efficient management of energy resources, electric utilities, and tra