Flood is one of the greatest disasters against oilfield manufacture. This paper first describes the key problem of oilfield flood control system, established a large oilfield flood control intelligent decision support system. Then the application research of reservoir grouppsilas flood prevention optimizing schedule algorithm was done based on multi-agent theory. According to the neural network mathematical model of oilfield reservoir, a novel algorithm of rainfall forecast was completed. Finally, on the basis of computer network technology and GSM wireless technologies, the application system was realized in order to expand the scope of flood dispatching decision-making method.