Method, device, system and storage medium for predicting physical variables of bow-cat

By introducing physically driven and data-driven physical Fourier neural operator networks and residual connection networks into the bow network system, the problem of insufficient robustness and generalization capabilities of traditional models in response prediction and state recognition of bow network system is solved, and a more accurate and flexible dynamic behavior prediction of bow network system is achieved.

CN119150706BActive Publication Date: 2025-03-18GUANGXI NANNING SHUANGZHUO TECHNOLOGY CO LTD
0 Cites 0 Cited by

Patent Information

Application Number
CN202411640775.5
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-11-18
Publication Date
2025-03-18
Estimated Expiration
2044-11-18

AI Technical Summary

Technical Problem

The traditional bow net interaction analysis model has poor robustness and generalization ability for the response prediction and state identification of bow net system, making it difficult to accurately characterize the dynamic behavior of bow net system.

Method used

A physically driven and data-driven physical Fourier neural operator network (PhyFnoNet) is proposed. The Fourier neural operator replaces integral operation, avoiding the need for integral discreteization, and a novel residual connection network is designed to establish a soft constraint relationship using the prior knowledge between physical variables.

Benefits of technology

It significantly enhances the performance of the model in solving the prediction of space-time dynamics in complex PCS environments, can adapt to different arch network parameters and conditions, and has strong ability to learn physical coupled equation systems.

✦ Generated by Eureka AI based on patent content.
Patent Text Reader

Abstract

The present application relates to the field of rail transit technology, and discloses a bow-net physical variable prediction method, device, system and storage medium, which are used to capture the complex mapping relationship between the internal parameters of the bow-net system dynamics model from the frequency domain through Fourier neural operators, and simulate the complex correlation between the bow-net system parameters through residual connection networks to improve the prediction results of target neural operators. The bow-net physical variable prediction method includes: obtaining the bow-net system input parameters corresponding to the target prediction task; inputting the bow-net system input parameters into the first fully connected layer for conversion to obtain the initial time domain state; inputting the initial time domain state into the multiple target neural operators of the feature extraction layer for iterative frequency domain feature extraction to obtain the target time domain state; inputting the target time domain state into the second fully connected layer for conversion to obtain the initial bow-net physical variable set; inputting the initial bow-net physical variable set into the residual connection network layer for correction to obtain the target bow-net physical variable set.
Need to check novelty before this filing date? Find Prior Art