Gimballed Stable-Platform Seeker: Line-of-Sight Rate Estimation via Pixel-Plane Kalman Filtering and Gimbal-Mounted Gyro Fusion

Abstract

This paper presents a line-of-sight (LOS) rate estimation method for a gimballed, stableplatform electro-optical seeker. Unlike a strapdown seeker, where the camera is rigidly fixed to the vehicle body and the body’s own rotation must be subtracted out of the measurement, a gimballed seeker carries the camera on an actively steered platform whose gyro senses the platform’s own inertial rate directly. This removes the need for a body-rate correction term, at the cost of having to model and compensate for the finite bandwidth of the gimbal servo itself. The method combines a pixel-plane Kalman filter, which tracks the target’s normalized image coordinates and their rates, with the gimbal-mounted gyro rate through the transport theorem, producing a three-component inertial LOS rate vector free of az/el singularities. The approach is tested on a single, deliberately severe scenario: the interceptor body executes an independent roll/pitch/yaw disturbance while translating along a sharpcornered path, and the target simultaneously traces a square path in azimuth and elevation relative to the body, with a finite gimbal servo time constant and realistic pixel-tracking noise. Under this scenario the estimator achieves a steady-state root-mean-square (RMS) LOS-rate error of 1.02 deg/s and an overall RMS error of 2.29 deg/s (max 12.2 deg/s), with the largest errors concentrated at the sharp corners of the target’s square path, where the LOS rate itself changes discontinuously.