The invention discloses a big data acquisition method, which comprises the following steps of S1, uniformly processing different types of data based on a visual process arrangement platform; S11, constructing a design state platform; S12, customizing different access components for different types of data; S13, presetting an access assembly; S2, uniformly arranging data reconciliation and data quality detection public flows based on a visual flow arrangement platform; S21, arranging a data reconciliation assembly and a data reconciliation process at a node capable of actively acquiring sourcedata and destination data; and S22, arranging a data quality detection assembly and process at a node capable of actively pulling data. According to the invention, the difference of data source typesis shielded by a pre-acquisition processor, and then the data can be processed by a unified real-time stream engine, so that the processing logic is relatively clear and simple; and meanwhile, publicprocesses such as data reconciliation and data quality detection are conveniently and uniformly arranged, and the quality and efficiency of the whole system are improved.