Study On Optimization Method Of Waterflooding Reservoir Based On Optimal Control And SPSA Algorithm

Water flooding is an important mode in secondary oil exploitation. During the flooding process, the injected water preferentially penetrates into the layer with high permeability due to the heterogeneity of reservoir geology. In this case, other routes are shorted by high permeability routes amid thus the oil in other region are difficult to be displaced and extracted. It is an urgent problem for the urgent need to improve the oil field in the water flooding and to improve the oil stable oil effect and slow the production decline of oil field.Basing on this background, The water flooding reservoir optimization control algorithm all exist the optimization effect was not obvious, this article proposes a optimal injection strategy and carries on recovery on flooding reservoir optimization algorithm. In reality, the reservoir development has a certain sense of direction and application.In terms of the dynamic system, the economic net present value is selected as objective function, the saturation is selected as state variables and oil production and water injection are regarded as control variables. Meanwhile, the constraint conditions are based on the continuity equations and boundary conditions so as to establish the mathematical model of the water flooding reservoir optimization.According to the reservoir characteristics, one is the gradient algorithm of optimal control algorithm. About the optimal choice of the control algorithm, this paper applies the steepest descent method, in this algorithm; the gradient of the solution is a major difficulty. Based on the ad-joint method of solving the gradient, to address the calculation speed is slow and complex shortcomings,this paper improves the ad-joint method to increase the computational efficiency and reduce computational complexity, thus forming a modified steepest descent method. The other is no gradient algorithm called SPSA (Simultaneous Perturbation Stochastic Approximation) algorithm. This algorithm is mainly through the approximation of the gradient of the objective function to approach the optimal solution, thus avoiding the complexity of the solution for the gradient of complex function. According to the process of solving the two algorithms, this paper design of water flooding reservoir optimization simulation program respectively, this program uses Eclipse set up reservoir model, and Programming simulation using MATLAB. The optimization method used in this paper can get better effect in actual reservoir optimization, and the cost is low, the human cost less.