Research ArticleMultilevel adaptive control of nonlinear interconnected systems

- A multilevel approach with global boundedness of states and output asymptotic tracking.
- Nonlinear controller as the supervisory controller for neutralizing the interaction.
- Adaptive backstepping-based design for the local controllers of the subsystems.
- Optimizing the control system using genetic algorithm.
- Modeling coupled inverted pendulums connected by springs, and simulation of method.

This paper presents an adaptive backstepping-based multilevel approach for the first time to control nonlinear interconnected systems with unknown parameters. The system consists of a nonlinear controller at the first level to neutralize the interaction terms, and some adaptive controllers at the second level, in which the gains are optimally tuned using genetic algorithm. The presented scheme can be used in systems with strong couplings where completely ignoring the interactions leads to problems in performance or stability. In order to test the suitability of the method, two case studies are provided: the uncertain double and triple coupled inverted pendulums connected by springs with unknown parameters. The simulation results show that the method is capable of controlling the system effectively, in both regulation and tracking tasks.