Login| Sign Up| Help| Contact|

Patent Searching and Data


Title:
METHOD AND SYSTEM FOR MULTI-STEP PREDICTION OF FUTURE WIND SPEED BASED ON AUTOMATIC RESERVOIR NEURAL NETWORK
Document Type and Number:
WIPO Patent Application WO/2022/033258
Kind Code:
A1
Abstract:
A method and system for multi-step prediction of a future wind speed based on an automatic reservoir neural network, which method and system realize accurate and fast multi-step prediction of future information, maintain relatively high robustness with regard to noise and a system time variation, and avoid the problem of over-fitting. The technical solution thereof involves: for short-term high-dimensional wind speed data, on the basis of a delay embedding theory, using a spatial-temporal information transformation to take observed high-dimensional dynamics as a reservoir, and mapping the high-dimensional wind speed data to future information of a target variable; and an automatic reservoir neural network realizing multi-step prediction of the target variable by means of solving a pair of conjugate spatial-temporal information interaction equations.

Inventors:
CHEN PEI (CN)
LIU RUI (CN)
CHEN LUONAN (CN)
Application Number:
PCT/CN2021/105931
Publication Date:
February 17, 2022
Filing Date:
July 13, 2021
Export Citation:
Click for automatic bibliography generation   Help
Assignee:
CENTER FOR EXCELLENCE IN MOLECULAR CELL SCIENCE CHINESE ACAD OF SCIENCES (CN)
CHEN PEI (CN)
LIU RUI (CN)
International Classes:
G06Q10/04
Domestic Patent References:
WO2013033284A12013-03-07
Foreign References:
CN107895202A2018-04-10
CN111242377A2020-06-05
CN107292447A2017-10-24
CN108615089A2018-10-02
US20200063710A12020-02-27
Attorney, Agent or Firm:
SHANGHAI PATENT & TRADEMARK LAW OFFICE, LLC (CN)
Download PDF: