The application provides a cargo positioning and anti-
collision system and method for the floating loading and unloading process of a semi-submersible
barge. By fusing the ship body
pose obtained by a ship basic
perception unit and the local
pose and
point cloud data of the cargo obtained by a
cargo scanning and tracking unit, a high-precision real-time three-dimensional
pose of the cargo in a unified coordinate
system is calculated by a data fusion module. At the same time, the
point cloud data is used to quickly generate a three-dimensional model of the cargo with the assistance of
artificial intelligence, and the model and the physical world cargo are driven to move synchronously in a digital twin environment, realizing the
visualization of the whole process of the floating loading and unloading operation. On this basis, the anti-collision
simulation module performs forward-looking motion trajectory deduction and interference detection according to the real-time pose of the cargo, transforming the traditional experience-dependent extensive operation into data and model-driven precise control, and significantly improving the cargo positioning accuracy,
operation safety and intelligent level.