A resonance suppression method in platform style inertial reference unit via particle swarm optimization notch filter

With the development of laser communication, remote sensing imaging, and other technologies, an inertial reference unit (IRU) plays an essential part in the line-of-sight (LOS) stabilization system used for acquiring, pointing, and tracking targets. The IRU provides a stable reference beam to realize accurate LOS pointing under external disturbances. Compared with the frame style IRU, the platform style IRU (PIRU) can achieve a higher bandwidth and better precision. However, mechanical resonance is introduced by a flexure hinge inevitably in the PIRU, which affects the performance of the LOS stabilization system. In this paper, an open-loop dynamic model of PIRU is established. Identification experiments are carried out with results indicating a 28.7 dB resonance peak at 27.07 Hz in the x axis and a 30.3 dB resonance peak at 26.59 Hz in the y axis. An asymmetric notch filter is used to suppress the resonance peak to achieve a higher control bandwidth. A fitness function is designed to represent the effect of resonance suppression. A particle swarm optimization algorithm is used to search for an optimal solution of the fitness function to obtain the parameters of the asymmetric notch filter. Experimental results show that the resonance peak is reduced by 97.88% and the system bandwidth reaches 159.31 Hz.