ABSTRACT
Soft-sensors are usually used for difficult-to-measure variables prediction in wastewater treatment. Nowadays, neural networks, especially the dynamic neural networks, are widely used to predict and monitor these variables. However, traditional training methods of dynamic neural network cannot fully consider the uncertainties of measurements and model, resulting in inaccurate estimations of parameters and deviation of prediction performance. Therefore, this paper proposes a novel adaptive multi-output soft-sensor. In the proposed method, the square root unscented Kalman filter (SR-UKF) is adopted to timely update the weights of neural network in each layer. Through this strategy, we are able to effectively improve the prediction accuracy of the standard neural networks. What's more, providing a new way to update the parameters online as the sequential property of Kalman filter. The proposed method is verified by a data set from University of California database (UCI database). The results illustrated that the proposed model can achieve better prediction accuracy compared with the traditional models.
- Huang D. P., Liu Y. Q., Li Y. Soft sensor research and its application in wastewater treatment. CIESC Journal, 62, 01 (2011), 7--15.Google Scholar
- Capel-Cuevas, S., Cuéllar, M., Orbe-Payá, I., Pegalajar, M. C. and Capitán-Vallvey, L. Full-range optical pH sensor array based on neural networks. Microchemical Journal, 97, 2 (2011), 225--233.Google ScholarCross Ref
- Yang Q. D., Wang F. L., Chang Y. Q. Soft sensor of biomass based on improved BP neural network. Control and Decision, 023, 008 (2008), 869--873.Google Scholar
- Guan, X. Z., Song, T. L., Yan-Hai, X. U. and Liu, Y. Comparison of Prediction on BP Neural Networks and Elman Neural Networks in Wastewater Treatment. Techniques of Automation and Applications (2014).Google Scholar
- Liang, Y. Application of Elman Neural Network in Short-Term Load Forecasting. City, 2010.Google Scholar
- Yang M. Y., Zhou F. Q., Li J. Soft sensing model for city wastewater treatment process parameter based on Elman neural network. Journal of Southeast University (Natural Science Edition), S1 (2006), 119--123.Google Scholar
- Guo X. Y., Guo M, Han H. G. Soft Measurement of SVI Based on PSO-BP Neural Network. Control Engineering of China, 21, 6 (2014).Google Scholar
- Pineda and Fernando, J. Recurrent Backpropagation and the Dynamical Approach to Adaptive Neural Computation. Neural Computation, 59, 2 (1989), 161--172.Google Scholar
- Williams, R. J. A learning algorithm for continually running fully recurrent networks. Neural Comput, 1 (1989).Google Scholar
- Kalman, R. E. A New Approach To Linear Filtering and Prediction Problems. Journal of Basic Engineering, 82D (1960), 35--45.Google ScholarCross Ref
- Snyder, C. Introduction to the Kalman Filter (2010).Google Scholar
- Julier, S. J., Uhlmann, J. K. and Durrant-Whyte, H. F. A new approach for filtering nonlinear systems. City, 1995.Google Scholar
- Merwe, R. and Wan, E. Sigma-Point Kalman Filters for Probabilistic Inference in Dynamic State-Space Models. City, 2003.Google Scholar
- Julier, S. J. The scaled unscented transformation. City, 2002.Google Scholar
- Merwe, R. and Wan, E. A. The square-root unscented Kalman filter for state and parameter-estimation. City, 2002.Google Scholar
- Elman, J. L. Finding Structure in Time. Cognitive Science, 14, 2 (1990), 179--211.Google ScholarCross Ref
- Werbos and P., J. Backpropagation through time: what it does and how to do it. Proc IEEE, 78, 10 (1990), 1550--1560.Google ScholarCross Ref
- Hagan, M. T., Demuth, H. B. and Beale, M. H. Neural network design. Neural network design, 2002.Google Scholar
- Hagan, M. T. and Menhaj, M. B. Training feedforward networks with the Marquardt algorithm. IEEE Transactions on Neural Networks, 5, 6 (2002), 989--993.Google Scholar
- Dua, D. and Graff, C. UCI Machine Learning Repository. City, 2017.Google Scholar
- Liu, Y., Huang, D. and Li, Y. Development of Interval Soft Sensors Using Enhanced Just-in-Time Learning and Inductive Confidence Predictor. Industrial & Engineering Chemistry Research, 51, 8 (2012), 3356--3367.Google Scholar
- Wu J. Study on Kernel Function Modeling Method for Unsteady Process of Wastewater Treatment. South China University of Technology, 2020.Google Scholar
Index Terms
- Development of Adaptive Soft-Sensors Based on Kalman-Elman Neural Network with Applications in Wastewater Treatment
Recommendations
Identification and control using MLP, Elman, NARXSP and radial basis function networks: a comparative analysis
This paper describes four neural networks multilayer perceptron (MLP) network, Elman network, NARXSP network and radial basis function (RBF) network. Neural networks are applied for identification and control of DC servo motor and benchmark nonlinear ...
Application of Elman Neural Network in Short-Term Load Forecasting
AICI '10: Proceedings of the 2010 International Conference on Artificial Intelligence and Computational Intelligence - Volume 02An Elman neural network with a weather component is proposed for the power load forecasting. Elman neural network can meet nonlinear recognition and process predition of the dynamic system, and make power system having the ability to adapt to time-...
Application of Neural Network Model in Insulating Oil Fault Diagnosis
ICMTMA '10: Proceedings of the 2010 International Conference on Measuring Technology and Mechatronics Automation - Volume 01:This paper introduced the practical value of artificial neural network in fault diagnosis field, respectively, applied the BP network, ELMAN network, RBF neural networks insulating oil fault diagnosis. Simulation results show that, in the transformer ...
Comments