An Application Of An Extremum-Seeking Algorithm With Improved Gradient Estimation For Propeller Noise Control

Abstract

Distributed propulsion systems are emerging in future electric aircraft concepts, often in the form of propeller powerplants. One of the major environmental effects that propellers present is their acoustic signature, which is especially concerning in the proximity of airports. In this paper, a model-free adaptive control approach is presented in order to minimize the tonal noise generated by the propellers of an aircraft with a distributed propeller propulsion system on specific ground locations. This method is implemented in the form of an extremum-seeking algorithm using an Extended Kalman filter as a gradient estimator. The underlying physical mechanism of the noise reduction manifests itself through destructive interference between the sound sources represented by the distributed propulsion system. Control of the location of the destructive interference spatial regions is achieved through controlling of the relative phase angles of the propellers. Feedback is achieved by an array of receivers located on the ground where a noise reduction is desired. A scalar cost function is evaluated on each measurement station based on a given sampling time to obtain a score of the perceived sound levels, to be minimized by the control algorithm. Example simulation cases are presented with and without active control, and the potential noise reduction is quantified.