Vector object-based architecture modeling method and condition to dense well data

Object-based reservoir modeling technology is often used to predict channel sand in river sedimentary systems. The disadvantage of this method is that it is difficult to obey the conditional data under dense well data. An optimization method based on vector-object is proposed. Conditional algorithms can utilize a cross-objective function that defines the matching degree of each object. Thousands of conditional objects are generated and stored in the database. Use three channel selection strategies to get the final implementation, and respect all well observation data, and sand proportions. Vector object-based methods are a combination of spatial vectors and objects. Geological objects are further abstracted as vector objects based on object-based. Using vector space analysis algorithms, it is easy to locally adjust geological objects to match hard data. This paper uses the open-source GIS computational geometry library GEOS (https://trac.osgeo.org/geos/) for vector computation, which contains very complete spatial queries and a large part of spatial operations. The test experiment proves that the modeling method based on vector-object is flexible and efficient, including high-curvature channel and non-stationary fan-channel simulations, and the proposed method can easily match dense well data.