Abstract: Urban ground collapses pose significant threats to human life and property safety.To identify the influence indicators of ground collapse, find out risk areas, and reduce potential disaster losses, a ground collapse risk assessment method was established.Taking Hengyang City as an example, the basic data of the drainage network were collected.The artificial neural network(ANN) was used to predict the size of the sewer breakage, and the logistic regression algorithm was used to predict the occurrence rate of ground collapse around the sewer.The results show that the average variance between the ANN predicted values and the true values is0.026,and the ground collapse risk is highly correlated with the sewer breakage according to the logistic regression algorithm.The Chengxi and Linhu drainage areas have the high risk of ground collapse, which might be caused by sewer breakage, surface loading, extreme rainfall, high-speed flow, etc.This study provides support for the prevention and control of ground collapse in urban areas of the middle reaches of the Changjiang River.