Spatial distribution prediction of landslide susceptibility based on integrated particle swarm optimization

Landslide sensitivity prediction relies on multiple environmental factors, making it difficult to obtain accurate prediction results. In order to improve the prediction accuracy of regional landslide sensitivity, a landslide sensitivity spatial distribution prediction method based on integrated particle swarm optimization was studied in Lianhe Village, Jianfeng Town, Shizhong District, Leshan City, Sichuan Province. Based on the determination coefficient, the sensitivity of landslide influencing factors was analyzed, and the weights of the influencing factors were determined. A landslide sensitivity spatial distribution prediction model was established based on support vector machine. By introducing simulated annealing and mutation operations into the particle swarm algorithm, an integrated particle swarm algorithm was obtained to extract high weight features of landslide sensitivity space and generate landslide sensitivity prediction results. The experimental results show that the cumulative value (ACU) of this method for predicting landslide sensitivity is 0.91, which can accurately predict the spatial distribution of landslide sensitivity in the study area and has practical value.