This invention discloses an intelligent stratification
system and method for static
cone penetration test curves based on a visual large-
scale model, belonging to the field of intelligent
processing and analysis technology for
geotechnical engineering investigation data. The
system includes a static
cone penetration test data acquisition and
standardization preprocessing module, which acquires multi-dimensional static
cone penetration test data such as depth, cone tip resistance, and sidewall friction resistance, performs preprocessing, and generates standardized static cone
penetration test curve images; a visual large-
scale model module, which constructs a visual large-
scale model to analyze the static cone
penetration test curve images, identifies key points, and outputs soil layer numbers,
soil properties, layer depths, and reasons for layering; a mathematical analysis module, which obtains layer depths based on a change point detection
algorithm and constructs a soil property naming rule
library, outputting layer and soil property naming results; and a fusion and output module, which performs comprehensive decision-making and naming arbitration, outputting layer results and soil property names. This invention achieves intelligent fusion of multi-
source data,
automation of layer identification, and structured result output, providing a new direction for the identification and classification of soil
layers in static cone penetration tests.