The present invention relates to an intelligent adaptive system and method for monitoring leakage of oil pipeline networks based on big data. The present invention effectively analyzes a large amount of data collected on site within a reasonable time period and obtains a state of a pipeline network by an intelligent adaptive method, thereby obtaining a topological structure of a pipeline network. The present invention specifically adopts a flow balance method in combination with information conformance theory to analyze whether the pipeline network has leakage; small amount of leakage and slow leakage can be perfectly and accurately alarmed upon detection; as a generalized regression neural network is adopted to locate a leakage of the pipeline network, an accuracy of a result is increased. Therefore, the present invention adopts a policy and intelligent adaptive method based on big data to solve problems of detecting and locating leakage of the pipeline network.