Adaptive Control of a Missile Fuel Control Valve

Adaptive control methods are investigated for a missile fuel control valve actuator system. The actuator possesses uncertain, nonlinear dynamics which include various torque bias components. A full state feedback design is proposed subject to a finite difference approximation of actuator velocity. Radial basis functions and feedforward neural networks are employed to estimate the uncertainty. The controller provides bounded tracking and guarantees a uniform ultimate boundedness of all the signals in the corresponding closed-loop system. An output feedback solution is also developed using an adaptive observer where the nonlinearity is represented as a polynomial with uncertain coefficients. The closed loop, output feedback system provides asymptotic tracking performance with bounded coefficient estimation. Simulations are conducted to evaluate and compare both designs.