I enjoy making things. Here are a selection of projects that I have worked on over the years.
This project aims to build model mainly uses the vision fusion transformer (ViFT) structure based on transformer to extract the spatial information of meteorological ground elements under different spatial resolutions, and adds text, mainly the longitude and latitude information of target sites or sections, as a cue word.
By connecting geospatial information with AI neural network structure, that is, integrating geospatial information on the basis of traditional neural network, the model can better understand and process spatial data information
Two classical model structures are employed to explore how different coupling strategies using optical flow method affect prediction results along with corresponding attribution analysis.
This project aims to optimize SWMM (Storm Water Management Model) parameters by Genetic Algorithm (GA) to simulate and predict urban runoff under rainfall events.
