Efficient Evaluation of Value of Information for Geotechnical Pre-posterior Analysis with BUS
Geotechnical structure monitoring can be designed and implemented to obtain insights into geotechnical structure responses.However,benefits gained from the measurements depend on the choice of monitored variables.Pre-posterior analysis provides a rational tool to support decision on the choice of monitored variables for risk-based optimization of geotechnical structure monitoring design.As for the pre-posterior analysis,one computational challenging task lies in estimation of failure probabilities for various monitoring scenarios with different possible monitoring outcomes so as to evaluate the value of monitoring information under a probabilistic framework.This paper proposes a Bayesian approach based on a recently developed Bayesian updating framework,namely Bayesian updating with structural reliability methods(BUS),to efficiently estimate and update the failure probability of the geotechnical structure concerned under various monitoring scenarios with different monitoring outcomes.The proposed approach is illustrated through a levee example.Results show that the proposed approach properly estimates the failure probability of uplift based on different monitoring outcomes.It does not need to re-run the calculation model to update geotechnical parameters as different possible monitoring outcomes are considered.This significantly improves the computational efficiency for pre-posterior analysis when the calculation model is complex and computationally demanding.
Geotechnical structure monitoring Pre-posterior Analysis BUS Value of information
Fu-Ping Zhang Zi-Jun Cao Dian-Qing Li
State Key Laboratory of Water Resources and Hydropower Engineering Science,Institute of Engineering Risk and Disaster Prevention,Wuhan University,Wuhan,Peoples Republic of China
国际会议
The 7th World Conference on Structural Control and Monitoring(7WCSCM)(第七届结构控制与监测世界大会)
青岛
英文
2206-2214
2018-07-22（万方平台首次上网日期，不代表论文的发表时间）