An implementation of blind-tendency behavior for covert human tracking robot

Covert tracking robot refers to the robot that can prevent itself from being found by the objective by hiding-in-mask or blind-tendency while it tracks a human or others, which can be applied into fields of public security, surveillance, military, criminals tracking, and et al. In this paper, an implementation method for the robot´s blind-tendency behavior is proposed based on geometric analysis, which is mainly composed by six sub-actions: RA, DL, DTCL, DTd0, DTH and EFH. Sub-actions RA can be used to adjust the robot´s attitude; by sub-actions DL, DTCL and DTd0, the robot can shorten its distance to the target to a predefined value while keeping certain blind-tendency ability under different cases; by sub-action DTH, the robot can slide to the central line of the human´s blind area while maintaining distance unchanged; sub-action EFH is designed to help the robot to escape from the occasional embarrassment of being too close to the human. Design of these sub-actions takes the robot´s motion smoothness and its nonholonomic constraint into consideration adequately. The method´s applicability, reliability, stability and robustness are confirmed by simulations with observing noises.