A Hybrid Genetic Algorithm for Optimal Control Solving of Polymer Flooding

This paper researches the optimization of injection strategies of polymer flooding in oil recovery. An optimal control problem (OCP) of a distributed parameter system (DPS) is formulated, in which the functional of performance index is profit maximum and the governing equations are the fluid equations in porous media. The control variables are chosen as the polymer concentrations and the slug size. The constraint conditions include boundary constraints and other inequality constraints. By a control vector parameterization (CVP) method, the OCP is transformed into a mixed integer optimization problem (MIOP). A hybrid genetic algorithm (HGA), which incorporates a position displacement strategy of the particle swarm optimizer (PSO) along with a special truncation procedure for handling integer restrictions, is applied to solve the MIOP. Finally, an example of the OCP for polymer flooding is exposed and the results show that the HGA method is effective and feasible.