Performance Evaluation of a Distributed Hybrid MPC Strategy Applied to the Four-Tank System *

摘要

The interest in control for distributed systems has led to the development of strategies that aim to overcome issues related to high computational burden or lack of robustness. Furthermore, the search for more convenient ways of modeling MIMO systems has motivated the development of techniques for the representation and control of hybrid systems that include both continuous and discrete variables.In this article we compare the robustness of a Distributed Hybrid MPC strategy against a Centralized MPC strategy, applied to the four-tank system. Performance was evaluated by testing random failure in sensors and actuators. Results show that for medium levels of failure, the distributed strategy has a better performance, but it requires more calculation time. The methodology proposed can be used in different applications, however it does not have assured convergence properties.