DocumentCode :
232001
Title :
Research on the flow control strategy of water distributor in Water Injection Well
Author :
Jia Deli ; Pei Xiaohan ; Liu He ; Sun Fuchao
Author_Institution :
China Res. Inst. of oil exploration & Dev.(RIPED), Beijing, China
fYear :
2014
fDate :
28-30 July 2014
Firstpage :
4993
Lastpage :
4996
Abstract :
The technological process of the water distributor flow deployment has mixed features of strong non-linearity, tough interference of multilayer coupling and complex models in Water Injection Well. The traditional model-based control strategies cause many problems (e.g. system instability, high time-consuming deployment, pressure built up in pipeline), which make it difficult to meet the expected requirements of technological process. In this paper, a novel PID-NN control strategy is proposed to solve the problems. The controller´s performance has been optimized by the strategy in three aspects as follows: modeling non-linear-controlled object, decoupling parameters of flow and pressure and improving efficiency and accuracy of the controller. The simulation results and pilot experiments of test well illustrate that this strategy improves the robustness and response speed of the system, and also proves this controller can handle multivariate decoupling problem of the technological process of the water distributor flow deployment.
Keywords :
drilling (geotechnical); flow control; neurocontrollers; nonlinear control systems; three-term control; PID-NN control strategy; controller accuracy; controller efficiency; flow control strategy; flow decoupling parameters; model-based control strategies; neural network control; nonlinear-controlled object; proportional-integral-derivative control; water distributor flow deployment; water injection well; Accuracy; Biological neural networks; Neurons; Plugs; Process control; Robustness; Measurement; PIN Neural Network; Water Distributor; Water Injection Well;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Control Conference (CCC), 2014 33rd Chinese
Conference_Location :
Nanjing
Type :
conf
DOI :
10.1109/ChiCC.2014.6895787
Filename :
6895787
Link To Document :
بازگشت