Multi-station runoff process stochastic simulation method and system based on improved variational auto-encoder

By improving the variational autoencoder model, the spatiotemporal distribution characteristics are directly learned from the three-dimensional runoff data, and high-quality multi-station runoff process simulation data is generated, which solves the problems of low simulation accuracy and efficiency in the existing technology, and achieves more efficient and more accurate random simulation of runoff process.

CN119830727AActive Publication Date: 2025-04-15HUAZHONG UNIV OF SCI & TECH
6 Cites 0 Cited by

Patent Information

Application Number
CN202411891809.8
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-12-20
Publication Date
2025-04-15
Estimated Expiration
2044-12-20

Smart Images

  • Figure CN119830727A_ABST
    Figure CN119830727A_ABST
Patent Text Reader

Abstract

The invention discloses a multi-station runoff process stochastic simulation method and system based on an improved variational auto-encoder. The method comprises the following steps: acquiring three-dimensional runoff data of a hydrometric station; the three-dimensional runoff data is utilized to train a variational auto-encoder model, the variational auto-encoder model comprises an encoder and a decoder, the encoder is used for converting the input three-dimensional runoff data into hidden space data, and the decoder is used for converting the hidden space data into simulation data; in the training process, the error between the simulation data and the input three-dimensional runoff data and the divergence between the input three-dimensional runoff data and the hidden space data are calculated, and when the error and the divergence are minimum, the corresponding parameters of the variational auto-encoder model are finally determined parameters; and by taking three-dimensional runoff data of a hydrometric station as input, generating new simulation data by utilizing the trained variational auto-encoder model, so as to realize random simulation of a runoff process. According to the invention, the problems of poor precision and low efficiency in runoff stochastic simulation are solved.
Need to check novelty before this filing date? Find Prior Art

Citation Information

Patent Citations

  • Runoff prediction method based on local attention enhancement model

    CN116050652A

  • River runoff prediction method and device based on mixed deep learning

    CN117114190A

  • Model training method, runoff prediction method, electronic equipment and storage medium

    CN118114057A

  • Blurred image general enhancement method based on hidden space diffusion model

    CN118154468A

  • DEM abnormal value detection method based on auto-encoder

    CN119006892A