This paper studies the optimal control problem for discrete-time nonlinear systems and an approximate dynamic programming-based Model Predictive Control (MPC) scheme is proposed for minimizing a quadratic performance measure. In the proposed approach, the value function is approximated as a quadratic function for which the parametric matrix is computed using a switched system approximate of the nonlinear system. The approach is modified further using a multi-stage scheme to improve the control accuracy and an extension to incorporate state constraints. The MPC scheme is validated experimentally on a multi-tank system which is modeled as a third-order nonlinear system. The experimental results show the proposed MPC scheme results in significantly lesser online computation compared to the Nonlinear MPC scheme.
@article{arxiv.2312.05952,
title = {Approximate Dynamic Programming based Model Predictive Control of Nonlinear systems},
author = {Keerthi Chacko and Midhun T. Augustine and S. Janardhanan and Deepak U. Patil and I. N. Kar},
journal= {arXiv preprint arXiv:2312.05952},
year = {2023}
}