Friction-induced sightline jitter significantly degrades the image resolution and detection range of a stabilised optical system. Therefore, any controller that can reduce jitter levels in the stabilisation sub-system will have a significant impact on overall electro-optic system performance. This article presents the results of an investigation into several friction compensation models applied to the validated model of an in-service electro-optic turret. The turret test harness, system identification software and friction measurement techniques used in the validation procedure-experimental transfer function analysis-are presented. A worstcase approach was used in setting the sensor noise and base motion acceleration levels. This test data was then used to validate a mathematical model of the turret elevation axis for use in off-line design and tuning of six friction compensation controllers. Three types of friction compensator model were investigated; a linear Kalman filter, an extended Kalman filter with a static friction model and an extended Kalman filter using a dynamic friction model. Additionally, two controller architectures were used. All six controllers were shown to significantly reduce jitter levels overall, but a new controller architecture was shown to also further reduce image degradation due to smearing.
This paper presents an extension to the original Frenet-Serret and Bishop frame target models used in the invariant extended Kalman filter (IEKF) to account for tangential accelerations for highly-manoeuvrable targets. State error propagation matrices are derived for both IEKFs and used to build the accelerating Frenet-Serret (FSa-LIEKF) and Bishop (Ba-LIEKF) algorithms. The filters are compared to the original Frenet-Serret and Bishop algorithms in a tracking scenario featuring a target performing a series of complex manoeuvres. The accelerating forms of the LIEKF are shown to improve velocity estimation during non-constant velocity trajectory segments at the expense of increased noise during simpler manoeuvres.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.