Explicit NMPC Based on Neural Network Models
This chapter considers an approximate mp-NLP approach to explicit solution of deterministic NMPC problems for constrained nonlinear systems described by neural network NARX models. The approach builds an orthogonal search tree structure of the regressor space partition and consists in constructing a piecewise linear (PWL) approximation to the optimal control sequence. A dual-mode control strategy is proposed in order to achieve an offset-free closed-loop response in the presence of bounded disturbances and/or model errors. It consists in using the explicit NMPC (based on NARX model) when the output variable is far from the origin and applying an LQR in a neighborhood of the origin. The LQR design is based on an augmented linear ARX model which takes into account the integral regulation error. The approximate mp-NLP approach and the dual-mode approach are applied to design an explicit output-feedback NMPC for regulation of a pH maintaining system.
KeywordsNeural Network Model Linear Quadratic Regulator Gaussian Process Model Recede Horizon Control Nonlinear Model Predictive Control
Unable to display preview. Download preview PDF.
- 5.Grancharova, A., Kocijan, J., Johansen, T.A.: Dual-mode explicit output-feedback predictive control based on neural network models. In: Proceedings of the 8th IFAC Symposium on Nonlinear Control Systems (NOLCOS 2010), Bologna, Italy, pp. 545–550 (2010), http://www.IFAC-PapersOnLine.net
- 8.He, X., Asada, H.: A new method for identifying orders of input-output models for nonlinear dynamic systems. In: Proceedings of the American Control Conference, San Francisco, CA, pp. 2520–2523 (1993)Google Scholar
- 17.Ogata, K.: Discrete-time Control Systems. Prentice Hall, Englewood Cliffs (1995)Google Scholar
- 19.Principe, J.C., Euliano, N.R., Lefebvre, W.C.: Neural and Adaptive systems. John Wiley & Sons, New York (2000)Google Scholar
- 21.Venkatasubramanian, V., McAvoy, T.J. (eds.): Neural network applications in chemical engineering. Special Issue of Computers & Chemical Engineering 16(4) (1992)Google Scholar