The Application of Thermodynamic Parameter Model Based on RGSA-RBFNN in Vacuum Resurgence Control Process
- DOI
- 10.2991/caai-18.2018.1How to use a DOI?
- Keywords
- vacuum resurgence; thermodynamic parameter model; RGSA; GSA; RBFNN
- Abstract
In traditional vacuum resurgence control process, there are some problems such as insufficient moisture regain, unsatisfactory moisture content absorption and large steam loss. In order to solve these problems, we put forward the vacuum pre-conditioner’s thermodynamic parameter model to control the vacuum resurgence process quantitatively. At first we proposed an Reinforcement Gravitational Search Algorithm (RGSA) to optimize the RBF neural network parameters, and then used the RGSA-RBFNN, GSA-RBFNN, RBFNN to establish the corresponding thermodynamic parameter model of the vacuum pre-conditioner. At last we used test data set to verify the model established by those three kinds of neural network. The results showed that RGSA-RBFNN has very strong function mapping ability, higher precision and more advantages than GSA-RBFNN or RBFNN. Using the model to control the process of vacuum resurgence, the moisture content of tobacco strips increases from 8% to 15%, which is 4 percentage points higher than 11% of the unmodeled one. The moisture content of tobacco leaf has been improved greatly. It has certain guiding significance for tobacco strips production of vacuum pre-conditioner.
- Copyright
- © 2018, the Authors. Published by Atlantis Press.
- Open Access
- This is an open access article distributed under the CC BY-NC license (http://creativecommons.org/licenses/by-nc/4.0/).
Cite this article
TY - CONF AU - Xinye Li AU - Yekuan Luo AU - Jiajing Wang AU - Yushan Hao PY - 2018/08 DA - 2018/08 TI - The Application of Thermodynamic Parameter Model Based on RGSA-RBFNN in Vacuum Resurgence Control Process BT - Proceedings of the 2018 3rd International Conference on Control, Automation and Artificial Intelligence (CAAI 2018) PB - Atlantis Press SP - 1 EP - 5 SN - 2589-4919 UR - https://doi.org/10.2991/caai-18.2018.1 DO - 10.2991/caai-18.2018.1 ID - Li2018/08 ER -