Development of evaluation algorithms for operator training system

This paper focuses on a new method of evaluating operator's plant manipulation capability by using an operator-training simulator. We developed task evaluation algorithm (TEA) to analyze and evaluate the operator's training results automatically after he or she finished the training by using graphic user interface (GUI) emulation of DCS. TEA includes two main modules, sequence evaluation module (SEM) and quantity evaluation module (QEM). SEM indicates whether the operator follows the correct operating sequence that is prepared in advance. On the other hand; QEM qualitatively specifies the deviation of each operator's task from the standard operating procedure. In this study, we applied the two algorithms into the network version of Yonsei operator training system (yOTS) that has been developed since 1994. Current yOTS contains various models of chemical processes such as batch reaction, batch distillation, ammonia synthesis and syn-gas manufacturing process. TEA-applied yOTS analyzes the trainee's operating capability more precisely and automatically. (C) 2000 Elsevier Science Ltd.