Reservoir group outlet structure failure assessment method based on Bayesian network

Abstract

The invention relates to a reservoir group outlet structure failure assessment method based on the Bayesian network. The reservoir group outlet structure failure assessment method includes the steps of establishing a Bayesian network topological structure, determining respective probability, calculating spillway system failure probability, and calculating the integral failure probability. The Bayesian network for failure probability determination is combined with a conventional event tree for determining stepped reservoir group spillway facility failure, and calculating the probability of downstream control step failure caused by upstream dam bursting flood is added as needed to form a complete Bayesian spillway facility failure assessment network. A reservoir group outlet structure failure is a complex multivariable system, and the assessment system can be used to analyze common cause failure and conduct backward inference. The assessment system can scientifically determine the value of information and determine whether new information needs to be acquired, so that the outlet structure failure analysis is gradually improved and becomes more scientific.

Description

technical field [0001] The invention relates to a Bayesian network-based method for evaluating the failure of drainage structures of reservoir groups, which is a hydraulic analysis and calculation method, a risk assessment of drainage structures and a risk prevention and control analysis of cascade reservoir group dam systems method. Background technique [0002] The frequency of extreme climates in the world is accelerating, and the concept of the maximum possible flood PMF or the maximum precipitation PMP of the reservoir dam can no longer meet the requirements of the new standard, because many existing rivers or river systems in the basin are already very complex, such as reservoirs formed by cascade development Groups, the design of floods should also consider the floods caused by extreme disaster-causing factors such as dam breaks and extreme climates. Therefore, from the perspective of ensuring the safety of the dam, the discharge capacity of the flood discharge facil...

AI Technical Summary

Problems solved by technology

Based on traditional statistical analysis, the Monte Carlo method is suitable for describing the impact of unrelated events on the results, but it is difficult to describe the relationship between events

Event tree, fault tree and decision tree methods can better express the logical relationship between different levels and the degree of correlation between events, but for the multi-variable complex system of cascade reservoir group discharge structure failure, and the need to analyze the common cause failure and reverse reasoning system, the above method is difficult to analyze and express

Images