Risk Assessment And Diagnosis Model Of Oil Well Based On Bayesian Network

The risk control of the workplace of well not only depends on management levels, but also influenced by lots of uncertainties, and makes it difficult to predict and diagnose. In recent years, many researchers hope that by means of risk prediction research to enhance understanding of the well site complex situation. Among them, the researchers from different aspects and angles to establish the model, and according to the actual situation to focus on the factors analysis of the technology, human being and objects.But current methods are mainly used in oil production stage and risk design projects to predict and diagnose risk factors and have lacks in using for overall risk evaluation and diagnosis for the daily works.The main contributions of the thesis are as follows; one is to research the risk management methods of the type of qualitative and quantitative, especially involved applied to oil industry. It was mentioned that many of them were too static, incomprehensible and inapplicable for daily activities.Other is to analysis the theory of Bayesian network, parameter estimation and structure study. And the same time and get a conclusion about the characteristics of Bayesian networks:to manage all risk factors.This paper gets inspire from Bayesian networks point and the weakness in normal risk management methods and applied Bayesian networks to oil working place in the reality. Built a model based on oil working place. Solved the weakness when use this model. In that period, there is a problem comes out that is cannot to management daily activities. But get done when use the time series and the Bayesian networks together. This kind of method which using Bayesian networks concept built model solves the problems from normal ways and gets a methodological innovation. Of course add to traditional risk management methods. All by those are just to aim for a better actual practice value and. Getting a further illustrate that Bayesian networks has a higher efficiency in operational risk management.