Abstract
This paper deals with a study on face graph characteristics and application of the face graph to plant operation surveillance, presenting mathematical formula containing 21 parameters for drawing expressive face graphs. Visual sensitivity analysis experiments of face graph elements such as eyes, eyebrows, nose, mouth and so on, and visual discrimination power analysis experiments showed that eyes, eye direction and face direction are points which the human visual sense is highly sensitive, and that if one uses this face graph to express multidimensional data he will be able to detect a very slight change in the phase relationship among them.
Application of this face graph to the display of 11-dimensional time series data in a heat exchanger plant model made it clear that the face graph has an advantage for making the global states and general trends of plant operation comprehensible to human operators.
This paper concludes that the face graph presented here is useful, thus, for a man-machine interface in plant operation.
