With the acceleration of urbanization in recent years, urban space is becoming more and more crowded. In order to solve the problem of limited space faced by smart cities, people have begun to focus on the development and utilization of underground space. In many developing countries, the number of underground facilities is showing a rapid growth trend of these underground spaces at any time may be subject to seismic hazards, so it is necessary to pay attention to its safety management and maintenance work. In order to avoid any unexpected losses, it is necessary to conduct continuous risk assessment using risk assessment methods and report anomalies before they occur. This paper presents an overview of a risk assessment system used to assess the expected damage to buildings from a possible earthquake under uncertain conditions. In this paper, the uncertainty caused by insufficient understanding of the interdependence of various parameters is considered through fuzzy relation. In addition, fuzzy probability is used to express the uncertainty of earthquake disaster possibility in the form of probability distribution. The experimental results show that the maximum difference between the predicted damage rate and the actual damage rate is 0.16% and the minimum is 0 when the system evaluates high-risk areas.