Arquitetura para um ambiente de treinamento representado em realidade virtual.

Abstract

The use of virtual environments for operator training has become a routine in many domains. Its advantages include cost reduction and personalized training giving the operator the opportunity to progress in its own rhythm, and the possibility of being trained in the use of new machinery and technologies. The vast majority of those virtual environment simulating tools are program-based which makes difficulty its maintenance and functional evolution. The Human Interface laboratory (LIHM) has developed a simulator prototype that uses as a model based simulator engine. The model is built in Colored Petri Nets (CPN), and the entire simulator was built using freeware software. As a drawback, one of the software tools stopped supporting an important functionality that prevented the prototype from being operational. This work proposes a new architecture for the simulator, which was conceived for supporting the training of operators in a substation control room as well as to support the human error studies performed in the LIHM. This new architecture contemplates new requisites such as persistence and and the appropriate developing tools are proposed as part of this work. As a result of this work, a basic simulator configuration was built validating the proposed architecture. To build this initial prototype it was also developed a driver to interconnect the various modules proposed in the architecture.