Nonlinear Kalman filters for aircraft engine gas path health estimation with measurement uncertainty

This paper is concerned with nonlinear Kalman filtering approach to aircraft engine gas path analysis with measurement uncertainty. The uncertain measurements are characterized by time delay and packet dropout. The delay step of physical parameters occurs randomly, and its probability is regulated by a set of uncorrelated variables following Poisson distribution and uniform distribution. Packet dropout is caused as the data are not collected in time or data buffer overflows. The novel nonlinear Kalman filters (KFs) are developed using a multistep recursive estimation strategy with self-tuning buffer in the presence of gas path measurement uncertainty. The data buffers are introduced in the state estimator, the length of which is adaptive to the information loss level. The algorithms run recursively using the new arrival data and buffer position information. With a more effective arrangement of the collected measurements in real time, the better estimation accuracy of gas path health status is expected. Simulations involving abrupt fault and degradation datasets of aircraft engines were carried out to numerically evaluate and compare the performance of the improved nonlinear KFs with their existing KFs in the context of health estimation with time delay and packet dropout. The test results demonstrate that the proposed methodology not only reduces the computational time but also obtains a satisfactory accuracy for state estimation in the cases of engine gas path measurement uncertainty.