Performance Analysis of an Adaptive Nonlinear State Estimator for ARW of Onboard FOG

Abstract

AbstractThis paper explores the ability of an adaptive nonlinear state estimator to estimate the angle random walk (ARW) coefficient of onboard fiber optic gyro (FOG) in real time. To overcome the off-line estimation of traditional Allan variance method, some online estimation methods have been developed to estimate the ARW coefficient which is a dominant noise in five basic noise sources of onboard FOG. However, most of these online estimation methods are based on linear state space model and valid for static FOG rather than onboard FOG. Although the other methods use a nonlinear state space model to estimate the ARW coefficient of onboard FOG, the estimation results include add errors. Based on the drawbacks and requirements, an adaptive nonlinear state estimator reported by us in our previous work tries to estimate ARW coefficient of onboard FOG. The comparison of results of FOG on a moving car noise analyzed by adaptive nonlinear state estimator shows that the proposed approach is more suited to estimate the ARW coefficient than the existing ones.KeywordsAngle random walkOnboard FOGOnline estimationNonlinear state-space model