Study On Tuning Of Fractional-order PID Controller

With the rapid development of modern industry and science technology, industrial equipment has become increasingly complex, large-scale, and controlled-object often exhibit several characteristics, such as complex-nonlinear, multivariable, strong coupling, etc. At the same time, the requirements of industrial processes are also rising. But the vast majority of industrial process control systems are still based on the PID control as the primary means of control. However, PID control parameters tuning now are mostly based on experience by engineers. It’s not only time-consuming, but also unable to achieve a satisfied result. PID control is based on the model, but accurately model of industrial processes are often difficult to obtain. Therefore, the software development based on both the system identification and PID parameter tuning is essential.In this paper, using the least squares method and of the internal model control theory (IMC) which is based on all-pole approximation, univariate and multivariable system identification and PID parameter tuning are programming implementation. Eventually, PID tuning softwareis emulated in the DCS system for chemical process.First of all, the univariate least squares method, the multivariate step-test method, and the identification method of least squares for multivariable system are discussed. Based on the above methods, the relevant programs are developed, and simulated.Secondly, based on the IMC-theory, the all-pole approximation method is introduced into the IMC-PID controller parameter tuning and deduced general formula of IMC-PID control parameters tuning by McLaughlin for both first-order and second-order delay system, which makes programming simulation applications much easier. Also this approach is extended to multivariable systems, and programmed by C++language to achieve univariate and multivariate Internal Model PID parameter tuning. The program is validated by a first-order time lag model. The simulation curves show that this method can be easily applied to programming, and the control effect is good with fast response, strong robustness.Finally, the developed software of system identification which is based on the PID parameter tuning is applied in chemical process of ethylene cracking furnace, and simulated in a laboratory environment through DCS system. Offline simulation results proved that the software is effective and implemented.