The Oil Reservoir Damage Diagnosis Of Decision-making Support System Based On Intelligence Computing

With sustainable development of our country' s economy, energy supply is becoming tenser and oil demand grows day by day, petroleum industry is the main backbone of the economy development. The practice in oil field and the results of the theory study show the adverse effect of reservoir damage on oil-filed development, it directly influences the development of the petroleum enterprise and oil supply. How to diagnose the reservoir damage and protect effectively reservoir has become the critical problem of decision-making in oil-filed production management. Because the reasons of reservoir damages are highly complicated and uncertain and the accuracy of the traditional artificial analysis method is low, therefore introducing the intellectual decision support method, and setting up decision support system (DSS) that can enhance reservoir damage diagnosis decision-making efficiency become necessary.In this paper, the research indicates that the diagnosis and evaluation of reservoir damage are substantially to find the answer for complex manage decision problem, it needs scientific and high-efficient decision-making support tools. IDSS can store lots of history data, models and experts knowledge that was obtained beforehand and studied now. Model can be used to analyze and compute issues, and can simulate man' s thinking process to infer the answers of problems. DSS has lots of excellences, such as great memory capacity for knowledge, models and data, self-study, high work efficiency, etc. In this paper, the IDSS for reservoir damage diagnosis is established to analyze the types and reasons of reservoir damage using the expert knowledge stored. Choosing an excellent method of reservoirs stimulation through the improvement of predicting model can enhancean accuracy and decision-making efficiency of reservoir diagnosis and protection. It can directly provide a powerful mean for the scientific oil-field management decision-making. Meanwhile. Meanwhile it enriches the management science theory and method in practice and provides the benefit for science decision-making of management problem in another domain.In order to implement the IDSS for reservoir damage diagnosis, this paper have carried out key research to the decision-making support model based on NN, the reservoir damage evaluation method based on fuzzy reasoning network and intelligence compute based on reservoir damage mechanism and effect factor analysis, and have completed the design, integration and implement of IDSS for reservoir damage diagnosis.In the research field of reservoir damage evaluation method based on NN, the reservoir damage diagnosis model and DSS structure are established based on forward NN, and BP arithmetic based on genetic ultra-linearity is brought forward. The reservoir damage evaluation model based on NN has preferably learning mechanism and the self-adaption mechanism to the environment.In the research field of reservoir damage evaluate method based on fuzzy reasoning network, the fuzzy knowledge and the denotation method of fuzzy inference rule and the formalization description of fuzzy logical inference are put forward. The fuzzy neural network model gathering the expert knowledge, the fuzzy neural network model of ration testing data and decision-making support is set up.In the research field of decision-making support model based on intelligence compute, the decision-making support model based on NN, weighted fuzzy reasoning network, data mining, fuzzy integration judge and evolvement compute are set up for intelligent calculation.In the structure design and realization field of reservoir damage diagnosis decision-making support system, the reservoir damage diagnosis decision-making support system structure and every functional models based on intelligence compute are proposed. Introducing cooperative working mechanism, the integrative data collection and the citation of expert knowledge, reservoir damage evaluation can be achieved. Integrating the qualitative and quantitative information in the domain of petroleum engineering, geology, well logging, well testing and problems related with reservoir damage, the identification, evaluation, diagnosis for reservoir damage are conducted, reservoir damage is predicted and plan of reservoir damage stimulation is optimized.The research achievement in this paper has already been conducted in some oil production factory in Daqing oil field, and the obvious effect has been achieved.