A novel approach in designing PID controllers using closed-loop data

Abstract

Most methods to compute PID parameters are based on system models. The model should describe the system properties accurately. In order to construct a highly accurate model, it is necessary to identify system parameters using input/output data so that it satisfies the richness condition of input/output data. However, it is not suitable to add such input signals to the real systems in industries. Therefore, it is desirable to directly determine PID parameters based on operating data. In this paper, a PID controller design scheme which does not need any descriptive model is proposed, in which a suitable set of PID parameters is directly computed based on closed-loop data. According to the control scheme, if closed-loop data of the control system and some user-specified parameters included in the desired reference model are given, suitable set of PID parameters can be obtained. The effectiveness of the control scheme is evaluated by two numerical examples and applied in an injection molding process.