Observer-based fuzzy adaptive fault-tolerant nonlinear control for uncertain strict-feedback nonlinear systems with unknown control direction and its applications

Baraka Olivier Mushage,Jean Chamberlain Chedjou,Kyandoghere Kyamakya
DOI: https://doi.org/10.1007/s11071-017-3395-3
IF: 5.741
2017-02-14
Nonlinear Dynamics
Abstract:This paper proposes a new nonlinear control scheme incorporating a state observer, a fuzzy neural network (FNN) and a new Nussbaum function for strict-feedback nonlinear systems by considering several challenges. These challenges are external disturbances, uncertain dynamics, unmeasured states, constrained input, unknown control direction, singularity issue, and actuator’s faults of different types. The scheme uses approximations of the unknown system’s dynamics provided by the FNN, the system’s states variables estimation provided by a model-free high-gain observer, and the control direction provided by the Nussbaum function. Compared to existing schemes, in addition to the fact that the new scheme can tackle simultaneously all the aforementioned challenges with better tracking performances, it also cancels the assumption about the positive definiteness of the control gain function found in many works. Thus, the scheme suites for more applications as it can be applied in cases where the control gain can be either semi-negative/negative-definite, semi-positive/positive-definite. Furthermore, the knowledge of the bounds for uncertain dynamics, actuation faults, FNN approximation errors and external disturbances is not required as it is for many other schemes. The effectiveness of the scheme is illustrated by its successful application to three examples, which are the pitch angle control for a Boeing 747-100/200 represented by the ultimate approximate nonlinear longitudinal model over up-and-away flight regime, the trajectory tracking control of a one-link manipulator actuated by a brush DC (BDC) motor, and the position tracking control for an inverted pendulum.
engineering, mechanical,mechanics
What problem does this paper attempt to address?