Non Linear Model Based Predictive Controller of a Buck Boost Converter
| Authors: | Pomar Martin, Universidade Federal de Santa Catarina, Brazil
Elias Normey Rico Julio, Universidade Federal de Santa Catarina, Brazil
Camponogara Eduardo, Universidade Federal de Santa Catarina, Brazil |
|---|
| Topic: | 6.3 Power Plants and Power Systems |
|---|
| Session: | Power Plants and Power Systems Modelling and Control |
|---|
| Keywords: | Model Based Predictive Control, Non Linear Systems, Optimization, Sequential Quadratic Programming, Buck-Boost Converter |
|---|
Abstract
This paper presents a Non Linear Model Based Predictive Controller (NMBPC) of a buck-boost converter (BBC). The NMBPC uses a non-linear prediction of the system outputs based on a discretization of the average continuous model of the BBC, assuming the duty cycle as the controlled variable. A classical quadratic cost function J is minimized at each sample time using a Sequential Quadratic Programming (SQP) optimization algorithm that guarantees that the obtained control action gives a local optimal value of J. The tuning of the controller parameters is defined to obtain a compromise between performance and robustness. Simulation results in a wide range of the output voltage show that the proposed control strategy yields very fast time responses even under varying load situations.
