The Hybrid Approaches for Forecasting Real Time Multi-step-ahead Boiler Efficiency

We study how to optimize the boiler efficiency of a steam boiler which is the most important component in a fertilizer plant. In particular, we have proposed several methods for forecasting when the trend of the boiler efficiency is going down so that some control parameters of the boiler are adjusted to keep its efficiency stably. This is a challenging task since the boiler efficiency is a noisy time series data. In this paper, we propose two different methods for forecasting the boiler efficiency by multi-step-ahead (MSA) in real time. The first method, namely RTRL-RFNN, that applies a MSA reinforced real time learning algorithm for recurrent fuzzy neural networks (RFNNs). RTRL-RFNN repeatedly adjusts model parameters of RFNNs according to the latest observed values. The second method, namely SE-RFNN, is a hybrid of stochastic exploration and RFNNs. To demonstrate the performance of our methods we implement two proposed methods and an existent method called RFNN. Moreover, we illustrate the experimental results on the same dataset collected from Phu My Fertilizer Plant, Petro Vietnam Fertilizer and Chemical Corporation, Petro Vietnam Group, Vietnam. The experimental results show that three methods are appropriate to be employed for forecasting the real time MSA boiler efficiency and both proposed SE-RFNN and RTRL-RFNN outperform RFNN.

[1]  Smriti Srivastava,et al.  Chaotic time series prediction using combination of Hidden Markov Model and Neural Nets , 2010, 2010 International Conference on Computer Information Systems and Industrial Management Applications (CISIM).

[2]  Ali Cheknane,et al.  Forecasting hourly global solar radiation using hybrid k-means and nonlinear autoregressive neural network models , 2013 .

[3]  Chin-Teng Lin,et al.  A neural fuzzy control system with structure and parameter learning , 1995 .

[4]  Diyar Akay,et al.  Comparison of direct and iterative artificial neural network forecast approaches in multi-periodic time series forecasting , 2009, Expert Syst. Appl..

[5]  Madhumita Panda,et al.  A Hybrid Differential Evolution and Back-Propagation Algorithm for Feedforward Neural Network Training , 2013 .

[6]  Xuewei Li,et al.  Chaotic analysis of traffic time series , 2005 .

[7]  Richard S. Sutton,et al.  Generalization in Reinforcement Learning: Successful Examples Using Sparse Coarse Coding , 1995, NIPS.

[8]  M. E. Muller,et al.  A Note on the Generation of Random Normal Deviates , 1958 .

[9]  Bart De Schutter,et al.  A Comprehensive Survey of Multiagent Reinforcement Learning , 2008, IEEE Transactions on Systems, Man, and Cybernetics, Part C (Applications and Reviews).

[10]  Wei Xu,et al.  Prediction of Chaotic Time Series of RBF Neural Network Based on Particle Swarm Optimization , 2014, ECC.

[11]  Abhijit Gosavi,et al.  Reinforcement Learning: A Tutorial Survey and Recent Advances , 2009, INFORMS J. Comput..

[12]  Nan Jun,et al.  Combined Method of Chaotic Theory and Neural Networks for Water Quality Prediction , 2010 .

[13]  Richard S. Sutton,et al.  Introduction to Reinforcement Learning , 1998 .

[14]  Li-Chiu Chang,et al.  Reinforced recurrent neural networks for multi-step-ahead flood forecasts , 2013 .

[15]  Richard S. Sutton,et al.  Reinforcement Learning: An Introduction , 1998, IEEE Trans. Neural Networks.

[16]  Amaury Lendasse,et al.  Methodology for long-term prediction of time series , 2007, Neurocomputing.

[17]  Václav Snásel,et al.  Optimizing Boiler Efficiency by Data Mining Teciques: A Case Study , 2014, CONF-IRM.

[18]  Risto Miikkulainen,et al.  Efficient Reinforcement Learning Through Evolving Neural Network Topologies , 2002, GECCO.

[19]  Samarjit Kar,et al.  Applications of neuro fuzzy systems: A brief review and future outline , 2014, Appl. Soft Comput..

[20]  S. Sorooshian,et al.  Watershed rainfall forecasting using neuro-fuzzy networks with the assimilation of multi-sensor information , 2014 .