|
|
|
| Prediction of Cement fCaO Based on Particle Swarm Optimization and Continuous Deep Belief Network |
| LIU Bin1,ZHAO Peng-cheng1,GAO Wei2,SUN Chao2,LIU Hao-ran1 |
1.Information Science and Engineering College of Yanshan University, Hebei Province Key Laboratory of Special Optical Fiber and Optical Fiber Sensing, Qinhuangdao, Hebei 066004, China
2.Electrical Engineering College of Yanshan University, Qinhuangdao, Hebei 066004, China |
|
|
|
|
Abstract It is difficult to accurately detect the content of free calcium oxide in cement clinker in real time, thus a predictive model is proposed based on continuous deep belief network. Firstly, according to the cement clinker production process, the main variables in the firing system on the effect of the free calcium in the cement clinker are obtained. With the use of grey relational analysis theory, the grey relational degree is obtained, ignoring minor factors, and a set of auxiliary variables is established. Secondly, the cement clinker fCaO content model is built by using the particle swarm optimization algorithm to optimize the parameters of continuous deep belief network. Finally, compared with prediction effect of least squares support vector machine and back-propagation, the simulation results show that the model is of high precision and strong generalization ability.
|
|
Received: 15 February 2017
Published: 12 April 2018
|
|
|
|
Corresponding Authors:
Peng-Cheng ZHAO
E-mail: 15033569346@163.com
|
|
|
|
［1］Nawaz A, Julnipitawong P, Krammart P, et al. Effect and limitation of free lime content in cement-fly ash mixtures ［J］. Construction and Building Materials, 2016, 102: 515-530.
［2］王卓,苑明哲,王宏,等. 水泥熟料质量指标的软测量建模研究［J］. 化工自动化及仪表,2006,33(6):53-54.
［3］刘文光,王孝红,于宏亮,等. 基于LS-SVM的水泥熟料煅烧过程fCaO软测量研究［J］. 控制工程,2008,15(S2):85-88.
［4］王秀莲,孙旭晨,王卓,等. 基于局部PSO-LSSVM的水泥f-CaO测量方法研究［J］. 控制工程,2014,21(6):807-811.
［5］赵朋程,刘彬,孙超,等. 基于IQPSO优化ELM的熟料质量指标软测量研究［J］. 仪器仪表学报,2016,37(10):2243-2250.
［6］蒋妍妍,李洪林. 基于改进粒子群优化LSSVM的水泥熟料fCaO软测量［J］. 现代化工,2014,34(6):152-155.
［7］Li W T, Wang D H, Zhou X J, et al. An improved multi-source based soft sensor for measuring cement free lime content［J］. Information Sciences, 2015, 323(C): 94-105.
［8］Yuan J L, Tao H Z. Prediction of Free Lime Content in Cement Clinker Based on RBF Neural Network［J］. Journal of Wuhan University of Technology (Materials Science Edition), 2012, 27(1): 187-190.
［9］HintonG E, Salakhutdinov R R. Reducing the dimensionality of data with neural networks［J］. Science, 2006, 313(5786): 504-507.
［10］杨景明,陈伟明,车海军,等. 基于粒子群算法优化支持向量机的铝热连轧机轧制力预报［J］. 计量学报,2016,37(1):71-74.
［11］Hinton G E. Training products of experts by minimizing contrastive divergence［J］.Neural Computation, 2002, 14(8): 1771-1800.
［12］陆艺,张培培,王学影,等. 基于PSO-BP神经网络的关节臂式坐标测量机长度误差补偿［J］. 计量学报,2017,38(3):271-275.
［13］杨慎涛,刘文波. 基于粒子群的神经网络测试生成算法［J］. 计量学报,2015,36(2):197-201. |
|
|
|
2018, Vol. 39 
