A FEASIBLE STUDY OF REBUILDING KARST FLUXDATA BASED ON ANN MODEL

摘要

Multiplayer and multipolar characters of groundwater system in Northwest karsts area determine the nonlinearrnrelationship between input and output factors. Artificial Neural Network—ANN supplies a gap of traditional statistics that cannotrnbehave the relationship satisfyingly, which is similar to groundwater at principle and model establishment points. Remolding results ofrnANN is better than traditional regression analysis method through remolding the water data of groundwater in Dishu, Guanxirnprovince, therefore using ANN model to remold the flux data of groundwater system is feasible.