A virtual reality-based training environment using haptic interfaces.

摘要

Stability is a critical issue in haptic simulation. Unlike other conventional robotic manipulators, haptic devices inherently function in close proximity to humans. The unpredictable nature of human operator and the virtual environment are the main sources of instability in a stand alone haptic system. In the case of teleoperative haptic systems the criteria for stability is far more complicated than those studied for single-user systems. Stability and traceability in telehaptic systems are some of the main technical challenges in the field of shared virtual environments. The methods proposed in this thesis provide accurate 3D force reflecting telehaptic model to guarantee a stable teleoperation in a virtual environment that minimizes the tracking errors between the master and slave haptic devices.; A control strategy is developed and investigated for maintaining and improving on the stability margins and achievable performances for telehaptic system as a training tool intended for bilateral interaction with virtual objects. Specifically a recursive adaptive method is developed and applied in order to minimize the tracking error between the master and slave haptic devices. A detailed comparative evaluation of the proposed control and estimation strategies is presented to illustrate the performance of the resulting controlled telehaptic system.; The research presented here is based on an implementation environment consisting of master and slave haptic devices operating on a single workstation, which implies that improving the system response and performance in a network environment remains as another challenge and future work. The development of higher quality graphics accelerators, processors, high speed networks and data communications, specific protocols for 3D graphics and haptics world, specific real time algorithms and even the development of 3D modeling languages affect this next generation technology.