Research On Real-time Information Intelligent Analysis Model Based On Drilling Process

In order to deal with and use the uncertainties and fuzzy information generated in drilling process effectively and reasonably, a study on real-time information intelligent analysis model based on drilling process is launched, from two major aspects-representation and reasoning of uncertainty knowledge in drilling field. The main research contents and innovations are as follows:Firstly, the architecture and main functions of the real-time information intelligent drilling accident analysis model based on ontology and Bayesian Network is designed; and on the basis of previous research, the drilling field accident diagnosis ontology is improved further through Protege, While taking into account the randomness of drilling accidents and probability characteristics (uncertainty) of human subjective expectations, Bayesian Network is introduced for the probabilistic expansion of drilling accident intelligent diagnosis ontology; Secondly, in order to play the advantages of ontology and Bayesian Network in knowledge representation and uncertainty reasoning fully, structural transformation is proposed between the existing drilling accident intelligent diagnosis probabilistic ontology, and detailed elaboration is made under the guidance of the transformation rules; Thirdly, the model is developed through the Programming language Java and Myeclipse as a platform under the environment of Windows XP, and the seamless integration of ontology building tool Protege and Bayesian inference platforms Netica is realized through Myeclipse; Finally, Under laboratory conditions, according to the parameters collected from well sites, sticking trouble diagnostic analysis as examples, the probability of various sticking trouble may occur is calculated under certain conditions of evidence, and the types of sticking trouble are diagnosed and analyzed accurately. By the results, the feasibility and accuracy of the real-time information intelligent analysis model based on drilling process is verified preliminary.