Title :
Research on intelligent decision-making system for selecting development method of heavy oil reservoir
Author :
Zhang, Jianjun ; Shi, Junfeng ; Wang, Yahui ; Du, Fanglan
Author_Institution :
Res. Inst. of Pet. Exploration & Dev., CNPC, Beijing, China
Abstract :
Choosing the right development method is to determine whether or not the heavy oil reservoirs can development successfully. At present, the development method of heavy oil is decided by experts of heavy oil reservoir development. So, different experts may obtain different results, which can´t avoid the subjectivity brought by experts. To the problem, the paper established an intelligent decision-making system applied expert system and neural network to select the development method of heavy oil reservoir intelligently. Expert system play the role of summarizing and working up experiential knowledge of heavy oil reservoir development by knowledge base and the role of simulating experts´ thinking process by inference engine. In the knowledge of KB, the weight of the sensitivity parameters of the heavy oil reservoir that affect the selecting significantly can not be accurately man-made determined. The article introduced three lays of neural network system to obtain the trusty weight through training a large number of samples. The Intelligent decision-making system for selecting development method of heavy oil reservoir combined expert system and neural network not only make use of expert´s valuable knowledge effectively, but also avoided the subjectivity brought by experts. The result of proven examples showed that the intelligent decision-making system can select the development method of heavy oil reservoir rapidly and scientifically.
Keywords :
expert systems; hydrocarbon reservoirs; inference mechanisms; neural nets; production engineering computing; expert system; heavy oil reservoir; inference engine; intelligent decision-making system; neural network; Combustion; Economics; Petroleum; Production; Reservoirs; Viscosity; Heavy oil reservoir; development method; expert system; neural network;
Conference_Titel :
Computer-Aided Industrial Design & Conceptual Design (CAIDCD), 2010 IEEE 11th International Conference on
Conference_Location :
Yiwu
Print_ISBN :
978-1-4244-7973-3
DOI :
10.1109/CAIDCD.2010.5681923