Finite-time-prescribed performance-based adaptive command filtering control for MIMO nonlinear systems with unknown hysteresis

This article focuses on an adaptive neural network (NN) control problem for multi-input multi-output nonstrict-feedback nonlinear systems subject to unmeasurable states and actuator hysteresis. Firstly, a NN observer is proposed to obtain the unmeasurable states. In addition, the radial basis function neural networks are applied to online approximate the nonlinear terms. And then, a variable separation technique is utilized to solve the algebraic loop problem generated by the nonstrict-feedback structure and the complexity problem is also dealt with by using a command filter design technique, which is easy to reduce the repeated differentiations of virtual control signals. Meanwhile, to satisfy the practical engineering application, a finite-time performance function is implemented to make the tracking error can enter into the preassigned range within a given time. By using the proposed controller, the boundedness of all closed-loop signals is guaranteed and the tracking error can enter into the prescribed bounded range within a precise setting time. Finally, the preponderance and effectiveness of the developed controller are revealed by simulation examples.

Automated summary

This article addresses the issues of unmeasurable states and actuator hysteresis in multi-input multi-output nonstrict-feedback nonlinear systems. It proposes a neural network observer to estimate the unmeasurable states and uses radial basis function neural networks to approximate the nonlinear terms. A variable separation technique is employed to solve the algebraic loop problem and a command filter design technique reduces the complexity. A finite-time performance function is implemented to ensure the tracking error enters a preassigned range within a specified time. The effectiveness of the developed controller is demonstrated through simulation examples.