| dc.description.abstract | In the oil industry, it is very important to know the current status of drilling processes which can be obtained by analyzing the data from sensors on the drilling engines. The data which oil companies get is complicated, so, in order to analyse the data, it has to be processed first.
There are several methods of intelligent data analysis such as JESS, Petri Nets, R functions, Bayesian Networks and so on. Which of the reasoning techniques can be used to process the data and how to use it in a system are left for users to research and develop. To resolve the problem, upon the study of many other data-processing methods, this paper proposes several novel models for data processing step by step.To validate the effectiveness and the feasibility of the models, the author designs and implements several related systems to interface the reasoning techniques into the systems. The functions of every module in the system and the interrelations between them are achieved in the form of class and the core data structure is described in detail as well.
In chapter 2, the author first analyses the characteristics of reasoning technologies for identifying use. In chapter 3, the author chooses JESS as the reasoning technique to process and monitor the data. Based on the monitoring results from chapter 3 and Petri Net technique, chapter 4 developes another data processing system called ‟SUP system‟ and also analyses the performance of the system. The first 2 models are only used for single server, but when there is a lot of data need to be processed, multi servers are required. In order to solve this problem, the author also does some research on Rserve in a distributed environment in chapter 5.
The results prove that some models and systems are well developed and the reasoning techniques are well used in the systems, but some other reasoning techniques have limitations in the related models due to the reason of researching time and the author‟s knowledge. | en_US |
