Font Size: a A A

Modeling And Parameter Identification For Pneumatic Control Valve Stiction

Posted on:2011-09-07Degree:MasterType:Thesis
Country:ChinaCandidate:Y CongFull Text:PDF
GTID:2120330332476126Subject:Control Science and Engineering
Abstract/Summary:PDF Full Text Request
Pneumatic control valves are the most commonly used actuators in process industry for their high performance. However, they frequently suffer from problems, such as stiction, due to hot working environment or viscous fluids flowing through in long term. Stiction always causes oscillations in control loops which reduces the control loop performance.30% of the control loops are oscillatory due to control valve stiction. Thus, the modeling and compensation of valve stiction has drawn increasing concern from academics and researchers in recent years. The main contents contained in the thesis are as following:1. Based on the tests done by former researches and ISA standards, a set of new test parameters is added to check Choudhury model and two-layer binary tree model provided by Si-Lu Chen et al.in 2008. The results show that the Choudhury model does not fully comply with ISA standards, neither dose two-layer binary tree model on the new parameters test. After analyzing the two models and their logic problems, this thesis proposed improved method, which can make the two models consistent with the physical properties of the pneumatic actuator valve with stiction properties and ISA standards.2. The stiction parameter identification method is proposed by using the improved Choudhury model and the PSO algorithm, and the range of the parameters is discussed and optimized. The effectiveness of the improved identification method is demonstrated by simulation. The simulation results show the differences between Choudhury model, two-layer binary tree model and their improved models, and which one is better for using the stiction parameter identification.3. The improved models and stiction parameters identification method is used for analyzing 51 PID control loops in the delayed coking plant from a petrochemical company. The results showed that 4 control loops have stiction properties in the 51 control loops, and the effectiveness of these methods.4. A valve stiction detection and identification software is developed for the purpose of potential industrial application. There is an introduction about the framework of the software and the realization of each component. The virtual effect of the software is also included.At last, the summarizes, and the future research work are introduced.
Keywords/Search Tags:pneumatic control valve, stiction, ISA standards, parameter identification, process monitoring
PDF Full Text Request
Related items