Adaptive Tracking Differentiator with Feature Recognition for Signal Processing in High-Altitude Test Facilities

The ambient pressure is an important indicator for ensuring stable and efficient testing of the aeroengine in the high-altitude test (HAT) facilities. With the complex and quickly changing test environment and storage limitation of the underlying hardware, processing the pressure signals to access its filtering and differentiating signals to participate in designing feedback controller becomes difficult. The integration step of a differentiator algorithm, which is typically closely related to the sampling period, affects the performances of filtering and differentiating. In this work, a feature recognition based adaptive algorithm is proposed to enable the tracking differentiator (TD) by adaptively selecting an appropriate integration step in real-time. In particular, the sampled signals are transformed into images and analyzed by a proposed feature recognition algorithm. This algorithm can transform the real-time signals into an indice of system's dynamic characteristic. Simulation and experiment results show that the proposed adaptive TD algorithm can effectively improve the filtering and differentiating performances compared with the original TD, and meet the signal processing requirements of HAT facilities.