Data‐driven post‐earthquake rapid structural safety assessment

摘要: Earthquake prone cities are exposed to important societal and financial losses. An part of these losses stems from the inability use structures as shelters or for generating economic activity after event of an earthquake. The is not only due collapse damage; it also lack of knowledge about their safety state, which prohibits normal use. Because a diagnosis required for thousands structures, city-scale assessment requires solutions that economically sustainable and scalable. Data-driven algorithms supported by sensing technologies have potential solve this challenge. Several ambient vibration monitoring studies buildings, before earthquakes, shown the extent damage in building correlated with decrease natural frequency. However, observed worldwide data may be representative specific factors such construction type, quality, material, age, etc. In paper we propose framework able progressively learn relationship between frequency shift state small number buildings city inspected an earthquake, information predict uninspected but monitored buildings. The capacity proposed perform prognosis validated applying methodology to 1000 having simulated shifts states.