Markov chain risk transfer analysis method and system based on multi-model coupling

Through the Markov chain risk transmission analysis method coupled with multi-model, the problem of risk prediction dimension disaster and small sample prediction poor results in the existing technology are solved, and more accurate risk transmission analysis results are achieved.

CN118710066BActive Publication Date: 2025-03-25SOUTH CHINA INST OF ENVIRONMENTAL SCI MEP
1 Cites 0 Cited by

Patent Information

Application Number
CN202411189119.8
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-08-28
Publication Date
2025-03-25
Estimated Expiration
2044-08-28

AI Technical Summary

Technical Problem

The existing risk conduction analysis methods have problems in risk prediction, such as risk prediction dimension disaster and poor prediction effect under small sample conditions, and are limited by the lack of historical risk data and future engineering scheduling changes.

Method used

The Markov chain risk transmission analysis method based on multi-model coupling is adopted. By collecting flood data, the risk factor sequence and risk value sequence are constructed, key risk factors are screened, the joint distribution and risk prediction model is constructed, the dimensionality reduction process is used, the risk interval is divided, the risk space transfer probability matrix is ​​constructed, and the risk transmission mechanism is obtained.

Benefits of technology

It effectively overcomes the limitations of risk transmission research by lack of historical risk data and future engineering scheduling changes, improves the accuracy of risk transmission analysis results, and solves the problem of risk prediction dimension disaster and poor prediction results of small samples.

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

Abstract

The present invention discloses a Markov chain risk transfer analysis method and system based on multi-model coupling, collects flood data in the study area, extracts typical flood events, constructs risk factor sequences and risk value sequences, and screens out key risk factors; constructs joint distributions between key risk factors, and randomly samples massive flood events; uses maximum entropy projection pursuit method to reduce the dimension of key risk factors of typical floods to obtain comprehensive risk factors, constructs a risk prediction model based on the comprehensive risk factor sequence and risk value sequence, inputs massive flood events into the risk prediction model, and obtains corresponding massive random risk values; divides the risk value interval into low, medium, and high risk areas, and constructs a risk space transfer probability matrix based on massive random risk values ​​to obtain a risk transfer mechanism. The present invention effectively solves the problem of risk prediction dimensionality disaster in risk transfer analysis, and the problem that conventional risk prediction models have poor prediction effects under small sample conditions.
Need to check novelty before this filing date? Find Prior Art