Trajectory optimization by using EMD- and ICA-based processing method.

• EMD- and ICA-based method has been introduced to process the data from FOG. • The combined method shows improvement on trajectory optimization. • The combined method shows improvement on deviation reduction. • The combined method is powerful on processing obvious bumps and shocks. The angular velocity data of Fiber-Optic Gyro (FOG) are key signals to trajectory calculation for a FOG-based curve measuring system. The shocks from bumps on the route will contaminate the signal and finally drift the calculated trajectory. An optimized processing strategy on the contaminated signals is able to suppress the drifting with almost no trajectory distortion. In this paper, the signals from FOG have been processed by using median filtering (MF), Empirical Mode Decomposition (EMD) and the combined method of EMD and Independent Component Analysis (ICA). The experiments are performed under different conditions. The frequency spectrum and deviation study before and after processing have been investigated to compare the performance of the employed strategies. The combined EMD and ICA method has shown the best performance on trajectory optimization and error suppression with almost no distortion. [ABSTRACT FROM AUTHOR]