The 17th Shandong Provincial College Students Extracurricular Academic Science and technology Works Challenge Cup competition will be held at the end of May 2021. A team composed of Chen Guodong, Yin Guangliang, Zhong Chao, Wang Yanzhong, Zhang Jinding, Zhao Xinggang and Zhao Mengjie won the second prize of Shandong Province. The topic of this work stems from the urgent need for the layout of intelligent oilfield strategy, and takes the regulation and optimization of reservoir waterflooding development and production as the demand drive. Its purpose is to solve the key problem of real-time optimization of reservoir production. Based on the three key technical links of reservoir automatic history fitting, production optimization adjustment and fine injection and production control device in oilfield development scheme design, the complex nonlinear function approximation ability of machine learning is given full play, the data and seepage mechanism are fused in the framework of agent assisted neural network, and the optimization and intelligent algorithm are comprehensively used based on data driven and model driven. The theory and method of efficient optimization of intelligent oilfield production which integrates the whole production process are established, so as to construct the real-time optimization system of intelligent oilfield production. Aiming at the production problems of traditional reservoir history fitting, such as long time, complex well location design and injection and production relationship adjustment, and large error of fine water injection, the system integrates artificial intelligence into the main link of reservoir development and production design, carries out the research of new generation information technology from the theoretical basis, builds solutions based on practical oilfield engineering problems, and implements the engineering application. It can greatly shorten the historical fitting time of reservoir numerical simulation model and accelerate the well pattern optimization design and the scheme design of injection-production relationship control. The research content of the work is the combination of software and hardware. Aiming at the difficulty of accurately regulating the oil and water flow due to complex working conditions in the underground, a set of intelligent multi-layer and multi-stage flow control device of the underground hydraulic is developed to realize real-time adaptive detection and control. In addition, professors Zhang Kai, Yao Chuanjin and Zhang Liming won the title of Excellent Advisor.