Theory And Application Of Oil Field Development Adjustment Plan Optimization Based On Intelligent Computing

In the middle and later stage of oilfield development, the main problem that the oil company is confronted with at present and the future is how to improve the benefit of oil field, stable oil production and control water cut. The key to solve this problem is to develop a scientific and rational development plan. The important content of oilfield development plan is to optimize the adjustment plan. The aim is to maximize the benefit of oilfield development, to stable oil production and control water cut under various geological conditions. Because of the mutual restriction between targets and constraints, it is important theoretical and practical significance to use mathematical programming model and optimization technology to adjust the scheme optimization design. At present, the research of the oil field company planning adjustment problem is to optimize the oil production unit, water cut, workload of subordinate oil production units based on the production decline law. Parameters, variables and constraints of planning scheme and model is mostly established based on mean and certainty condition. The differences of geological structure and potential between different development blocks, injector-producer group and single well is not considered, and the impacts of the seasonal and artificial habits of the work are also not considered in detail. This leads to the larger differences of the actual situation and the planning objectives. Therefore, it is necessary to study the optimization model of oilfield development adjustment plan to guide the work of the oil production plant in the actual production process. The model involves the uncertainties, human factors, and geological conditions.The subject is carried out based on the typical problems in the optimization of oilfield development and adjustment. The multi-objective programming model is explored to establish by taking into consideration various factors, and using the intelligent computation theory to solve the related model. The content of this thesis is divided into 3 parts from the levels.(1) Research on Optimization Algorithm.Aim at the problem that the shuffled frog leaping algorithm is not high accuracy and easy to fall into local optimum, Combined with the advantages of cloud model theory, reverse learning theory, quantum computation and cultural algorithm, three kinds of improved shuffled frog leaping algorithm is proposed(Adaptive Grouping Chaotic Cloud Model Shuffled Frog Leaping Algorithm, Quantum Shuffled Frog Leaping Algorithm, Adaptive Mixed Culture Shuffled Frog Leaping Algorithm). The convergence of the algorithm is analyzed. (2) Research on classification model. The microstructure of the microstructure is identified by using the process neural network model and the log information, and model is trained by the AGCCM-SFLA. Finally, the single well and well groups are classified by fuzzy comprehensive evaluation theory. The membership function of the judging model is based on data distribution and cloud model, which is more consistent with the actual data and human thinking. (3) The establishment of nonlinear programming model.The wells of the dynamic and static inconsistencies are evaluated by the potential effect of the measure. Based on this, the multi-objective and multi-constraint comprehensive development and adjustment programming model and the stochastic programming model for oilfield development and adjustment are established. In the model, the different single wells, the construction month, the law of the output decline and the uncertainty of the oil price are considered. The model is solved by the AGCCM-SFLA and the QSFLA. At the same time, the schemes of the obtained programming model are evaluated by the multilevel fuzzy comprehensive evaluation method, the optimal working plan is given, and the model is verified through actual oil field development data, and achieve good practical results.For the typical application of oilfield development adjustment plan, this paper begins to identify the microscopic pore structure of reservoir by using the neural network and log data. The production status of single well and well group is classified by the dynamic and static data of single well and well group. The problematic wells are found in the current working state. It is to provide a collection of oil and water wells to be adjusted for the following adjustment plan. Providing the methods and models for oilfield development adjustment plan, which has great theoretical and application value for oilfield development and production.