Application of the improved Kalman filter algorithm to the servo tracking system

Abstract

In the boat satellite TV receiving system, the quartz gyro sensor is used to sense the rotation rate of all dimensions, and provide parameters for servo of posture. According to requirements of the system performance index, after analyzing the range of each direction's angular speed variation of the carrier and considering cost performance, in this task we use quartz gyro of CRS03.Angular velocity information measured by gyro scope can become angle information by integral transformation and become angular acceleration information by differential transformation. The noise in signal which is exported by quartz gyro sensor is very obvious even the quartz gyro sensor is at a stand still. If the signal is used to track directly or after simple filtering, tracking performance is poor even cannot tracking. Zero wander of random drift in gyro is anon-stationary random process. Before selecting appropriate filtering method we need modeling for gyro noise signal. The noise out put from gyro can be described by AR model in a sufficiently large order. According to the servo tracking system requirements, Kalman Filter Algorithm is used to forecast the target's position and speed. Consider processing tracking signal in case of the target is blocked, we improve Kalman Filter Algorithm. The system obtains good effecting exactly forecasting the position of target's image and the smooth of target's speed. In practical application, the system in high speed and high precision realized steady tracking. It can reach tactical and technical requirements of tracking system.