This paper proposes a BP neural network model based on genetic algorithm optimization (GA-BPNN) to predict the settlement of earth-rockfill dams. Based on the collected settlement monitoring data of an earth-rockfill dam in Xinjiang

the corresponding settlement points in the three-dimensional coordinates are obtained. The filling information of the upper and lower dam body at the monitoring points

 along with time parameters

is also input as training data for the neural network. The optimal fitness parameters calculated by the genetic optimization algorithm are used as the initial weights and thresholds for the neural network to improve its efficiency in finding the global optimal solution. The results indicate that the GA-BPNN is a viable method for predicting the settlement of earth-rockfill dams. The GA-BPNN demonstrates superior predictive performance compared to the BPNN

exhibits greater stability

 and yields results that are closer to the actual values.