论文部分内容阅读
提出一种有效的实时纵向飞行轨迹重构的新方法。为了得到状态估计的快速算法,本文把非线性飞行轨迹重构转化为线性、离散、时变状态和参数估计问题。将数值稳定性好、计算量也小的序列U-D分解滤波算法用于状态方程为线性、观测方程为线性或非线性的滤波问题中。由于测量值中常常含有系统偏差,本文把这些偏差作为增广状态加入增广状态模型中,并利用模型的一些特点,提出偏差分离的U-D分解算法,使计算量大大减少。仿真和实际试飞数据计算表明、本文的方法可得到比平方根协方差滤波更有效的实时飞行轨迹重构结果。
A new method of reconstruction of real-time longitudinal flight path is proposed. In order to get a fast algorithm of state estimation, this paper transforms nonlinear flight path reconstruction into linear, discrete, time-varying state and parameter estimation problems. The U-D decomposition filtering algorithm with good numerical stability and small amount of calculation is used in the filtering problem that the state equation is linear and the observation equation is linear or non-linear. Since the measured values often contain systematic deviations, we add these deviations as augmented states to the augmented state model. By using some characteristics of the model, a U-D decomposition algorithm with deviation separation is proposed, which greatly reduces the computational cost. Simulation and actual test data calculation show that the proposed method can obtain more effective real-time flight path reconstruction results than square root covariance filter.