Direct Error Observation Compensator for Visual Servo Tracking System With Image Sensor-Based Sparse Measurement

For a visual servo tracking system, a direct error observation (DEOB) compensator strategy is proposed to break the bandwidth limitation caused by sparse measurement of the image sensor. The essence of this method lies in reconstructing errors through inverting the controller for direct counteraction rather than indirectly reducing the controller output to eliminate the closed-loop error. Furthermore, to tradeoff the DEOB bandwidth and system stability, a dynamic feedforward path is conducive to obtain flexible tuning degrees of freedom in comparison with the traditional Youla-Kucera (Y-K) parametrization. Due to as a direct compensation method in addition to being an alternative implementation of Y-K parameterization, the control stability and dynamic response can be well balanced to achieve high performance. As a result, it is convenient for engineering implementation irrespective of signal coupling and independent of the accurate model. Simulation and experimental results demonstrate the effectiveness of the DEOB in a visual servo stabilization tracking platform.