Extreme Learning Machine Assisted Adaptive Control of a Quadrotor Helicopter

Abstract EN

Control of quadrotor helicopters is difficult because the problemis naturally nonlinear.

The problem becomes more challenging forcommon model based controllers when unpredictable uncertainties anddisturbances in physical control system are taken into account.

This paperproposes a novel intelligent controller design based on a fast onlinelearning method called extreme learning machine (ELM).

Our neuralcontroller does not require precise system modeling or prior knowledgeof disturbances and well approximates the dynamics of the quadrotorat a fast speed.

The proposed method also incorporates a sliding modecontroller for further elimination of external disturbances.

Simulationresults demonstrate that the proposed controller can reliably stabilize aquadrotor helicopter in both agitated attitude and position control tasks.