Design of a Data-Driven Performance-Adaptive PID Controller
Toru Yamamoto, Sirish Shah, Takayuki Kono, Takao Hinamoto
In industrial processes, in order to improve the productivity, it is necessary to establish the performance-driven control mechanism, which the control performance is firstly evaluated, and the controller is reconstructed.
Moreover, since most processes have nonlinearities, controller design schemes to deal with such systems are required. This paper describes a design scheme of data-
driven performance-adaptive PID controllers. According to the proposed control scheme, the data-driven system identification which is effective for nonlinearity
works corresponding to the result of modeling performance assessment, and PID parameters are computed using the newly estimated system parameters. Finally, a simulation example is discussed.