Model reference adaptive control of a class of non-autonomous systems using serial input neuron

The serial input neuron is an effective approximator of real-valued functions defined on the real line. In this paper, a serial input neuron-based model reference adaptive controller is proposed for the control of a class of uncertain non-autonomous systems. An on-line learning rule is developed to approximate the unknown nonlinear terms so that closed-loop stability as well as internal signal boundedness can be guaranteed using traditional Lyapunov design as long as the approximation error is sufficiently small. Simulation examples are provided to demonstrate the design of the proposed control system. In addition, an off-line learning simulation result is also presented for comparison.