Abstract:
The implementation of model predictive controllers on resource-constrained embedded platforms requires fast and computationally efficient processors to solve constrained ...Show MoreMetadata
Abstract:
The implementation of model predictive controllers on resource-constrained embedded platforms requires fast and computationally efficient processors to solve constrained convex optimization problems. We propose a set of design rules that make use of both, analytical guidelines and numerical simulations to achieve fixed-point implementations of an efficient interior-point (IP) algorithm for solving quadratic-programming (QP) problems. This is the first an necessary step in the development of a complete semiautomatic tool for the implemention of hardware accelerated predictive controllers for embedded and high throughput applications.
Published in: 2019 6th International Conference on Control, Decision and Information Technologies (CoDIT)
Date of Conference: 23-26 April 2019
Date Added to IEEE Xplore: 02 September 2019
ISBN Information: