Estimation of States of Aircrafts by Kalman Filtering Algorithms

摘要

The theory of optimal estimation has applications to a broad range of problem areas. Given the system model and noisy measurements, with known statistical properties of the noise in system dynamics and measurement devices, and initial condition information, an optimal state estimation algorithm obtains minimal variance estimates of the states of a linear system. State estimation techniques find extensive use in many areas: chemical processes, nuclear reactors, telecommunications, and aerospace vehicles. The problem of estimation of states of aircrafts is addressed using both real and simulated data. The data used for state estimation pertained to both short period and phugoid longitudinal modes of the aircraft. The state model used in the filter algorithm makes use of the parameters determined from the simulated/real data by maximum likelihood estimation (MLE) algorithm. The state estimation becomes necessary in case of postflight data analysis. This is because the acquired data are contaminated by random noise, measurement errors and biases. In this context, it was felt necessary to develop and utilize Kalman filter type algorithms to serve the need as starting and complementary algorithms for obtaining complete analysis of postflight data. A recent algorithm based on the concept of modifiable nonlinearity has been employed to handle real flight data, not only to filter the states (prefiltering mode) but also to estimate stability and control derivatives of an aircraft (complementary mode to MLE).