基于PSO-SVM模型的隧道水砂突涌量预测研究

复杂工程地质条件下,隧道水砂混合物突涌的预测防控是隧道安全建设的基础,准确预测水砂混合物突涌量,为工程提供安全保障至关重要。为提高预测准确性,提出一种基于粒子群算法优化的支持向量机(PSO-SVM)的隧道水砂突涌量预测模型。综合考虑地质构造、气象条件、施工影响三类因素,选取七个因子,结合某公路隧道,利用PSO-SVM建立隧道水砂突涌量预测模型,并对该隧道水砂突涌量进行预测,结果与实际突涌量一致。证实综合粒子群算法和支持向量机优势的PSO-SVM方法预测精度高,且易于实现,为类似隧道工程突涌预测提供参考与借鉴。

Study on prediction of water and sand inrush quantity in tunnel based on PSO-SVM model

The prediction and prevention of water and sand mixture inrush in tunnel under complicated geological conditions is the foundation of tunnel safety construction. Predicting the inrush quantity of water and sand mixture accurately is quite important for providing safety support to engineering. In order to improve the prediction accura-cy,the forecasting model for inrush quantity of water and sand mixture based on support vector machine combined with particle swarm algorithm optimization（ PSO-SVM）was presented. Taking a road tunnel as engineering back-ground,the geological structure,meteorological conditions and construction influence factors were selected as the major elements by considering of seven determiners,the forecasting model of tunnel water and sand inrushing was established based on PSO-SVM. The prediction process by the model was conducted and the well-pleasing results were acquired. The results showed that the comprehensive method can effectively improve the performance of pre-diction. Based on above conclusion,PSO-SVM is an approving method,and easily to be implemented,which pro-vides a significant technical mean for prediction of water and sand mixture inrush in tunnel,and presents notably useful reference value for engineering practice.