Abstract
In this paper, we present a novel model predictive control (MPC) scheme that incorporates stability information derived from a control Lyapunov function (CLF) to dynamically prune suboptimal sequences from the search space and decrease the computational burden placed on the controller. The CLF used for pruning is then incorporated into a cost function that penalizes energy in the error system as well as energy loss due to switching. Despite the very small control periods allowed due dynamic pruning, experimental results are given, showing the resulting controller generates low switching frequencies and low total harmonic distortion.

This publication has 36 references indexed in Scilit: