An iterative nonlinear predictive control algorithm based on linearisation and neural models

This paper is concerned with a computationally efficient suboptimal nonlinear predictive control algorithm. The nonlinear model of the plant is used to obtain a local linearisation and to calculate, by means of an iterative procedure, the nonlinear response and future control moves. In comparison with fully-fledged nonlinear algorithms, which hinge on non-convex optimisation, the presented approach is more reliable and less computationally demanding because it results in a series of convex, constrained or unconstrained, quadratic programming problems whereas its closed-loop performance is similar. The algorithm implementation for feedforward neural-network models is also discussed in the paper.