Block Dynamic Surface Control Applied to a Sea-Skimming Missile

Abstract

Nomenclature C Aσ= aerodynamic axial force coefficient derivative with respect to equivalent fins angle, rad −1 C lδ a = aerodynamic rolling moment coefficient derivative with respect to aileron angle, rad −1 C mδ e = aerodynamic pitch moment coefficient derivative with respect to elevator angle, rad −1 C Nδ e = aerodynamic normal force coefficient derivative with respect to elevator angle, rad −1 C nδ r = aerodynamic yaw moment coefficient derivative with respect to rudder angle, rad −1 c ref = chord of reference, m C Yδ r = aerodynamic lateral force coefficient derivative with respect to ruder angle, rad −1 F x ; F y ; F z = forces in body-axis direction, N h ref = reference height commanded, m I 0 = inertia tensor, kg ⋅ m 2 I 3×3 = 3 × 3 identity matrix M x ; M y ; M z = torque represented in body axis, N∕m m = mass, kg q = dynamic pressure, N∕m 2 S ref = reference surface, m 2 t = time, s V 0 = average velocity, m∕s v x ; v y ; v z = velocity in body-axis direction, m∕s x, y, z = Cartesian coordinates of position, m α, β = angles of attack and sideslip, rad α 0 = predicted angle of attack when in steady-state flight, rad δ a ; δ e ; δ r = aileron, elevator, and rudder equivalent angles, rad δ a 0 ; δ e 0 ; δ r 0 = predicted steady-state flight angles of aileron, elevator, and rudder, rad ϕ, θ, ψ = roll, pitch, and yaw angles, rad ω x ; ω y ; ω z = angular rate represented in body axis, rad/s

Robust missile autopilot design based on dynamic surface control

“…A controller signal interpolation strategy proposed by Lawrence et al was utilized to generate a group of linear autopilots [9]. The gap metric technique was introduced by Theodoulis and Duc to find the operation points [10]; it does not require the design of linear controllers for several points. Although this gain scheduling approach has been widely applied to missiles, it does not guarantee theoretical stability in the case of frequent and fast gain switching or in highly nonlinear conditions.…”

“…Another disadvantage of this method is that the gains are calculated a priori. Thus, any uncertainty not considered in the model can cause loss of performance or even instability in more extreme cases [10]. In addition, the necessary precise predictions of the aerodynamic coefficients over a wide range of flight conditions demand costly and timely wind-tunnel assays [11].…”