DRNN Based Decoupling Algorithm of Self-Tuning Controller on Line

2013 ◽  
Vol 442 ◽  
pp. 430-436
Author(s):  
Bo Bi ◽  
Xiao Yi Yang

In order to solve the puzzle that the change of a loop circuit parameter results in operation parameters change of other loop circuit in the control system, the paper proposed a sort of decoupling control algorithm of online self-tuning based on DRNN. In the paper, it took the temperature and humidity control of a certain controlled object as an example, constructed the mathematic model, analyzed the coupling relationship among the system variable, designed the decoupling network. It transforms the multi-variable control system with coupling relationship as the independent single-variable control system so as to eliminate the effect among related control channels. Based on decoupling algorithm of DRNN proposed in this paper, it made the research on system simulation experiment, and the response of system simulation demonstrated that it is very small to the effect of two channels of temperature and humidity control after through decoupling, and realized the decoupling among coupling variables. The results of simulation research show that the proposed decoupling control algorithm is feasible and reasonable.

2010 ◽  
Vol 43 ◽  
pp. 308-311 ◽  
Author(s):  
Rong Biao Zhang ◽  
Xian Lin Huang ◽  
Li Hong Wang ◽  
Jing Jing Guo

In view of characteristics of large time delay, multi-interference and strong coupling in temperature and humidity control system, an adaptive decoupling strategy based on generalized predictive control (GPC) and multi-model control is proposed in this paper. The proposed strategy mainly contains muti-model control, GPC decoupling control and adaptive algorithm. In multi-model control, multi-model sets are established to prevent the model mismatch in different working conditions. Meanwhile, this paper designs adaptive dynamic decoupling algorithms based on the principle of GPC. In actual experiments, temperature and humidity achieve precision of ±0.2°C and ±0.5% respectively.


Sign in / Sign up

Export Citation Format

Share Document