An Intelligent prediction model for UCG state based on dual-source LSTM
- PDF / 1,108,660 Bytes
- 10 Pages / 595.276 x 790.866 pts Page_size
- 10 Downloads / 172 Views
ORIGINAL ARTICLE
An Intelligent prediction model for UCG state based on dual‑source LSTM Yuteng Xiao1 · Hongsheng Yin1 · Tianhong Duan2 · Honggang Qi3 · Yudong Zhang4 · Alireza Jolfaei5 · Kaijian Xia1 Received: 3 March 2020 / Accepted: 20 September 2020 © Springer-Verlag GmbH Germany, part of Springer Nature 2020
Abstract Underground coal gasification (UCG) is a serious attempt to clean and efficient use of coal, but it has not been able to solve the problem of stable production. Predicting UCG can provide effective guidance for control, which effectively solves this problem. Existing UCG prediction models are not accurate, and most of them can only predict a single variable, and cannot adequately predict the UCG state. The paper proposes the concept of combustible gas equivalents that can characterize the concentration of mixed gas through stoichiometry and material balance equations. The equivalent gradient is introduced to characterize the trends in equivalent, and the UCG state discrimination standard is established to evaluate the UCG state. Eventually, a dual-source long short-term memory (LSTM) prediction model is proposed for predicting UCG state. The experimental results show that compared with Support Vector Machine (SVM) and Back Propagation Neural Network (BPNN) prediction model, the model can make a better prediction of equivalent value and the accuracy of predicting trends in equivalent reaches 90.99%. Keywords Equivalent · LSTM · Prediction · Underground coal gasification
1 Introduction UCG, one of the important development directions for clean and efficient utilization of coal [1–3], bypasses the traditional coal-burning process, has the advantages of low pollution emission [4], high resource utilization [5] and owns broad application prospects [6–8]. The buried coal is directly and incompletely combusted in-situ by injecting gasification * Hongsheng Yin [email protected] * Honggang Qi [email protected] 1
School of Information and Control Engineering, China University of Mining and Technology, Xuzhou 221116, China
2
School of Mines, China University of Mining and Technology, Xuzhou 221116, China
3
School of Computer Science and Technology, University of Chinese Academic of Sciences, Beijing 100049, China
4
Department of Informatics, University of Leicester, Leicester LE17RH, UK
5
Department of Computing, Macquarie University, Sydney, NSW 2109, Australia
agent, and after removing sulfide and nitrogen oxides, it eventually produces clean syngas with H2, CO, C H4 and other combustible gases as the main components [9]. Although UCG has many advantages, it has not been able to form a large-scale commodity development because of unstable gas production and calorific value, the low gasification rate, as well as the small and volatile combustible gas composition [10]. Gasification process was controlled by using the flow of the gasification agents mixture, the form of the gasification agents mixture and the pressure inside the ex-situ reactor [11]. Predicting the UCG state can g
Data Loading...
