Missile H_∞ gain scheduled autopilot design by resolving generalized eigenvalue problem

In this paper, generalized eigenvalue problem is constructed to obtain the optimal solution of H_∞ gain scheduled problem. By use of the method of sigular value decomposition, the linear matrix inequalities to solve H_∞ gain scheduled problem were transformed to a generalized eigenvalue problem in order to obtain optimal solution for the H_∞ gain scheduled problem. An H_∞ gain scheduled autopilot was designed for a surface-to-air missile by use of D-K-D iteration updated by the generalized eigenvalue problem. Simulation results show that the dynamic characteristics of the obtained autopilot are quite stable, when the flight conditions of missile are changing extensively. The results also demonstrate that the construction of the generalized eigenvalue problem is valid and the application of the generalized eigenvalue problem to H_∞ gain scheduled problem is feasible.