Identification and parameter sensitivity analyses of time-delay with single-fractional-pole systems under actuator rate limit effect

Abstract

A growing body of evidence has suggested that fractional-order models are able to describe dynamic processes with higher accuracy, such that thermal conduction processes are found to be depicted with time-delay with single-fractional-pole (TDSFP) models. Actuator rate limit phenomenon is extremely common in manufacturing due to either physical limitation or production safety guarantee. The popular step test signal in system identification has infinity derivative at step time which is not affordable by physics actuators. Thus numbers of industrial processes considered as traditional integer-order systems ignoring actuator rate limit have model mismatch to a greater or lesser extent. In this study, an identification method is proposed for TDSFP processes considering rate limit effects. Parameter sensitivity analyses are carried out to interpret the importance degree of each model parameter as well as the actuator rate limit value. Finally effectiveness of the proposed identification method considering the rate limit effect is validated on a hardware-in-the-loop real time temperature control system.