2001 年 37 巻 4 号 p. 199-207
This study proposes an adaptive skill learning aid system using an active controller as a training system for human-machine systems that require a high level of human skill for operation. The machine system (controlled system) is a tractor trailer system driving backward. It is simplified and linearized, and then its motion is simulated by a personal computer. Information about a manipulated variable (teaching manipulated variable) is forced to a trainee by the active controller along with the visual display of the motion of the machine system. The teaching manipulated variable and the visual display assist the trainee in understanding the causal (input-output) relationship of the machine system. Additionally, the teaching manipulated variable is adjusted so that it may adapt to an individual's control skill level during his/her learning process. This facilitates the acquirement of the targeted control skill and reduces the training load. Thus, the trainee learns the control skill for the position control of the machine system through the integrated observation of the visual information and forced information. Experimental verification of the proposed adaptive skill learning aid system indicated that the trainees could learn the control skill successfully within a relatively short training time.