1987 年 13 巻 4 号 p. 525-528
The control scheme recently proposed by the authors for diffusion processes has been extended so as to be usable for systems with unknown parameters. The extended Kalman filter algorithm was used for simultaneous estimation of parameters and state variables. The performance of estimator and controller was tested in real time in an experimental heat conduction process. The results show that the control scheme is robust and performs well even with few sensors.