A new method of missile in-flight alignment under large heading error

Aiming at large initial heading error at missile flight, a new in-flight alignment method is introduced, in which a H_∞ filter and a Kalman filter are synthesized. When there is large initial heading error, the H_∞ filter is used, at the same time, the observability of each state variable is calculated. When the observability of attitude angle error reaches a level prearranged, the Kalman filter with the covariance matrix inherited begins to work instead of H_∞ filter. The two filters are carried using the same error model which is designed under inertia coordinate system. The paper also introduces a method of calculating observabilities, which is not only adequate for the static observation equation case but also dynamic. At last, simulation result shows that, with large initial heading error, the precision of estimating attitude error can reach a level under 10´ within 120 seconds using the method introduced by this paper, while it maybe causes large estimating error and even causes filtering divergence using only H_∞ filter or Kalman filter.