Designing a Model Predictive Control system requires an accurate analysis of the interplay among three main components: the plant, the control algorithm, and the processor where the algorithm is executed. A main objective of this analysis is determining if the controller running on the chosen hardware meets the time requirements and response time of the plant. The constraints, in turn, should be met with a satisfactory tradeoff between algorithm complexity and processor performance. To carry out these analyses for an autonomous vehicle control, this paper proposes to leverage parallel co-simulation between the plant, the model predictive controller and the processor.
Co-simulation of a Model Predictive Control System for Automotive Applications
Bernardeschi C.;Dini P.;Domenici A.;Palmieri M.;Saponara S.;
2022-01-01
Abstract
Designing a Model Predictive Control system requires an accurate analysis of the interplay among three main components: the plant, the control algorithm, and the processor where the algorithm is executed. A main objective of this analysis is determining if the controller running on the chosen hardware meets the time requirements and response time of the plant. The constraints, in turn, should be met with a satisfactory tradeoff between algorithm complexity and processor performance. To carry out these analyses for an autonomous vehicle control, this paper proposes to leverage parallel co-simulation between the plant, the model predictive controller and the processor.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
