Premium
An edge computing‐based predictive evaluation scheme toward geological drilling data using long short‐term memory network
Author(s) -
Yin Junluo,
Luo Xiong,
Zhu Yueqin,
Wang Weiping,
Wang Long,
Huang Chao,
Wang JenqHaur
Publication year - 2021
Publication title -
transactions on emerging telecommunications technologies
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.366
H-Index - 47
ISSN - 2161-3915
DOI - 10.1002/ett.3888
Subject(s) - computer science , edge computing , field (mathematics) , enhanced data rates for gsm evolution , cloud computing , reliability (semiconductor) , scheme (mathematics) , data pre processing , data mining , term (time) , preprocessor , artificial intelligence , distributed computing , mathematical analysis , power (physics) , physics , mathematics , quantum mechanics , pure mathematics , operating system
Edge computing has been identified as one of the new computing paradigms, while improving service response performance and reliability in cloud computing environment. Meanwhile, there has been a tremendous surge of interest in the study of analyzing geological data. Given the success of edging computing in some industrial applications, it is also expected that it can be applied in this field, with the ever‐increasing demand for effectively achieving mineral exploration. Then, a novel application is accordingly developed in this article, through the combination of edge computing devices and geological data analysis model. Specifically, on the basis of multiple drilling data in the mine, a predictive evaluation scheme is proposed. After preprocessing the complex original data and analyzing the relationship between mineral composition content change and depth, the long short‐term memory (LSTM) as a recurrent neural network in the field of deep learning, is used to model certain mineral components, due to the satisfactory performance of LSTM in addressing time series data. Moreover, an evolutionary computing method, that is, the particle swarm optimization algorithm, is also incorporated into our scheme to optimize some key parameters, in an effort to further improve the computational performance. Finally, in edge computing environment our scheme is deployed on a typical edge device, that is, the Raspberry Pi, to achieve mineral exploration in a more efficient manner, while avoiding the limitation of resource concentration. The experimental results show that our edge computing‐based predictive evaluation scheme can effectively improve the performance in analyzing geological drilling data.