Development of a Simulation-Based Intelligent Tutoring System for Assisting PID Control Learning

Takeki NOGAMI  Yoshihide YOKOI  Ichiro YANAGISAWA  Shizuka MITUI  

Publication
IEICE TRANSACTIONS on Information and Systems   Vol.E77-D   No.1   pp.108-117
Publication Date: 1994/01/25
Online ISSN: 
DOI: 
Print ISSN: 0916-8532
Type of Manuscript: Special Section PAPER (Special Issue on Intelligent CAI and Hypermedia)
Category: 
Keyword: 
intelligent tutoring system,  simulation,  student model,  PID control,  system evaluation,  

Full Text: PDF>>
Buy this Article




Summary: 
A simulation-based ITS (Intelligent tutoring system), SRIM, has been developed for the purpose of providing individualized learning to students of PID control. We first indicate that the following two steps will be a burden to the student during personal use of simulators: 1) Selection of operational goals and 2) Interpretation of the simulation results. In order to reduce the burden of students in learning with a simulator, SRIM guides the learning process by providing local goals for PID controller tuning and by giving messages. Two tutoring strategies: i.e. the exercise style strategy and the illustrating style strategy, are employed in SRIM. In the exercise style strategy, a local goal for tuning a PID controller is first given to the student. A local goal is defined as one which can be satisfied by a single operation step such as Decrease the off-set." The student selects his operation and executes the simulation. By observing the simulation, the student understands whether his operation was a success or a failure. The illustrating style strategy is invoked to repair the student's erroneous knowledge when a contradiction is detected in the student model or a wrong operation is selected repeatedly. The architecture of ITS is employed to perform the local goal selection and the tutoring strategy switching, in a natural, well timed manner. The performance of SRIM was evaluated for the purpose of demonstrating the effectiveness of the teaching strategy. The evaluation experiment was carried out in the following steps: 1) Pre-test, 2) Learning and 3) Post-test. The teaching effect of SRIM was compared with other learning methods such as simple use of simulators or a textbook from the results of the pre-test and the post-test. The results showed that SRIM is effective in providing individualized learning with simulators.