Design of Output Feedback Controllers for Constrained Linear Systems Using Data Cluster Analysis
Amanda Danielle Oliveira da Silva Dantas1; André Felipe Oliveira de Azevedo Dantas2; Carlos Eduardo Trabuco Dórea1
1 Universidade Federal do Rio Grande do Norte; 2 Universidade Potiguar
Resumo
The design of output feedback controllers for linear systems subject to state and control constraints, additive disturbance and measurement noise, can be realized through the construction of an output-feedback controlled invariant set together with the multiparametric linear programming technique. This technique is commonly used to characterize the optimal control input as a piecewise affine (PWA) function over polyhedral regions. However, the multiparametric programming technique, when applied to multiple output systems, may require online high computational cost and high memory capacity. Within this context, this paper aims to use the K q-flat data cluster analysis algorithm to compute static output feedback control laws over a smaller number of polyhedral regions. An optimization problem is then proposed to compute sub-optimal output feedback control laws to further reduce the number of regions, as is illustrated by a numerical example.
Palavras-chave: Control; constrained linear system; invariant sets; multiparamentric programming; data cluster analysis