Design of Underground Powerhouse Cavern Support based on Particle Swarm Optimization Support Vector Machine
The support design of underground powerhouse caverns in hydropower stations faces major challenges due to the complexity of the underground environment.Existing solutions are limited by subjective experience and low accuracy,making it difficult to meet design require-ment.In order to improve the efficiency and reliability of underground powerhouse cavern design,an intelligent design model for underground powerhouse cavern support is developed by introducing particle swarm optimization(PSO)to optimize support vector machine(SVM)param-eters.The model uses the cavern span,cavern height,cavern height-to-span ratio,cavern burial depth,surrounding rock type,rock satu-rated uniaxial compressive strength,maximum principal stress value,and rock strength-to-stress ratio as input indicators.Through the train-ing and testing of 100 domestic and foreign underground hydropower station cavern support cases.The results show that the model shows a high degree of accuracy in various output indicators,among which the classification accuracy of spray-mix thickness,anchor diameter,and anchor spacing reaches 90%,85%,and 90%respectively,and the quantitative prediction goodness of fit of anchor length is 0.843.The study can provide a new method for underground powerhouse cavern support design.
国家科技期刊平台
NSTL
万方数据
