Deformation Prediction System of Concrete Dam Based on IVM-SCSO-RF

Deformation prediction is an important part of concrete dam safety monitoring. In recent years, the random forest (RF) algorithm has attracted more and attention in field monitoring because its fast speed strong generalization ability. However, performance RF easily affected by many factors, such as drift measured value displacement inappropriate setting parameters RF. To solve above problems, indicator variable model (IVM) used to identify eliminate values this paper, sand cat swarm optimization (SCSO) applied optimize for first time. On grounds this, a deformation system based on IVM optimized SCSO proposed. The case study shows that can correct interference data accurately, maximum error rate less than 3%; aspect parameter RF, results are obviously better those TAE method PSO algorithm, corresponding OOB minimum; terms performance, compared with TAE-RF, PSO-RF, LSTM SVM, SCSO-RF higher accuracy stronger stability, SSE MSE reduced at least 91%, MAE RMSE 71%, R2 very close 1. provide new automatic online evaluation performance.