Modeling, Optimization And Control Of Water-Injection Oil-Driving System In Oil Field

Water- injection is an important way to oil exploitation used by many oil-fields in our country, oil field water injection system is a complex fuild network. The energy consumption of oilfield injection system getting higher and higher, it is an urgent work to optimize the operation of the water injection system of oilfield. Combining the system's characteristic and the real requirements, this thesis made research of energy saving optimization of water-injection system,aimed at its modeling, restricted terms establishment, solution algorithm design, and result analysis.After analyzing and modeling the water-injection system in oil-fields,the operation scheme optimization of complex water injection system is established to find the optimized operation scheme based on fast simulated annealing. According to this scheme, the thesis designed the optimization solution algorithm. The simulation results show that the proposed method is effective and accurate enough.The efficiency of the water injection pumping station is a complex and multi-variable function, involving many elements which lead to great difficulty in strategy research of efficiency optimization. The factor which influences the efficiency of water-driving motor in oil field is complicated, so it is difficult to use precise mathematics model to describe the model of water-driving motor quantitatively. This paper proposes a high-order CMAC-type neural network——HCMAC, which is capable of implementing error-free approximations to multi-variable polynomial functions of arbitrary order. HCMAC Neural Network employed to design energy saving optimization control scheme and solve the optimization problem. In addition, to test the validity of the optimization algorithm, extensive simulation are conducted with some practical test data and the simulation results have proved its validity of the designed scheme.Dissolved oxygen (DO) is an important water quality parameter in the process of sewage treatment, which has great influence on the water quality and is not convenient for measuring directly. The factor which influences dissolved oxygen in the process of sewage treatment is complicated, so it is difficult to use precise mathematics model to describe dissolved oxygen quantitatively. This paper proposes a RBF neural network based soft measuring model of dissolved oxygen, which is employed to predict the dissolved oxygen. The simulation results show that the proposed method is effective and accurate enough for meeting the accuracy requirements for the design of sewage treatment feedback control system.Dissolved oxygen plays a crucial role during the design process of PID control strategy for sewage treatment. Dissolved oxygen control algorithm is developed based on first order plant model, and the parameter regions of the robust stability can be determined by using a graphical approach. Simulation results are provided to illustrate the design procedure and the effectiveness of the proposed methods.