The Research Of Intelligent Early Warning System For Drilling Based On Rule And Multi-source Information Fusion

Petroleum drilling engineering is a hidden underground project. The process of drilling has a lot of ambiguity, randomness and uncertainty problems, which requires rich experienced experts or technicians to real-time monitor. The number of experts and technicians who have a lot of experience is shortage, so people make use of drilling accident early warning system to make real-time early warning. The method of establishing the early warning model is very complex, and each kind of accident must be established itself early warning model. In view of the above reasons this paper proposes the intelligent early warning system for drilling based on rule and multi-source information fusion.This article describes a method. Firstly using the rough set theory of information fusion technology to build the early warning information system model, which can facilitate the knowledge base to establish and maintenance; Secondly, we research the algorithm of extracting rule and take advantage of the importance degree of condition attribute relative to decision attribute to replace the rule credibility of production rule, which solves the problem of needing to prior probability; Then research the early warning system in rule reasoning method combining with the characteristics of petroleum drilling; Last we design the drilling intelligent warning system completely and make it come true with the software. Through the inspection of the accident case, we can get the result that this theory has a strong action in rule extraction and accident inference. The system can integrate multi-source information and predict the accident type of drilling complex and accident in time. At the same time it can provide staff with the cause of the accident, the characteristics of the accident, the prevention and treatment measures and so on. Above all, this system is very meaningful for safe, rapid and efficient drilling.