基于粒子群算法的圆形隧洞锚喷支护优化

地下工程施工过程中,不同的支护方案具有不同的稳定性状态和经济成本,传统的列举择优法并不能得到所有可能方案的最优解。针对埋深大且围岩强度低的圆形隧洞,结合支持向量机与粒子群算法,对考虑支护时机的隧洞支护方案进行优化,得到最优方案,并以数值模拟进行对比验证。对比结果显示:采用本文所用优化方法得到的最优解的综合评价指标值与同方案数值模拟得到的综合评价指标值相对误差为0.137%,二者结果非常相近,且对比方案综合评价指标值大于最优方案的综合评价指标,说明算法寻优得到的结果可靠。通过本文研究,表明:(1)结合支持向量机与粒子群算法,可对地下工程支护方案进行全局优化;(2)优化结果经数值模拟验证有效;(3)优化方法与结果对相应工程设计及施工建设具有一定参考意义。

Bolt-shotcrete Support Optimization of Circular Tunnel Based on Particle Swarm Optimization

In the underground engineering construction process,different support schemes will bring different stability state and economic cost,traditional list preferred method can't get the optimal solution of all the possible solutions.Aiming at the deep buried and low intensity surrounding rock of circular tunnel,combining support vector machine and particle swarm optimization approach,tunnel support scheme was optimized considering supporting timing.The optimal solution was contrasted and verified by numerical simulation.Contrast results shown that:the relative error of the comprehensive evaluation index between the optimal solution obtained by the optimization method and the same scheme by numerical simulation is 0.137％,the results are very close,and the comprehensive evaluation index of the comparison scheme is larger than the optimal scheme.Through this study,it's proved that:(1) the method combined with SVM and PSO can optimize the underground engineering support scheme;(2) the optimization results are validated by numerical simulation;(3) the optimization method and results can provides reference for the design and construction of similar engineering.