基于并行遗传算法和温度荷载的隧洞参数反演研究

针对隧洞工程运行期反演计算中温度荷载容易被忽视的问题,以西南地区一个大型水电站的深埋地下隧洞为研究对象,将其运行期钢筋应力、围岩参数增量的实测值作为反演的基础数据,考虑应力场和渗流场的耦合作用,采用并行遗传算法对运行期隧洞围岩参数和外水头进行反演。计算结果和实测结果对比表明,温度荷载对反演得到的隧洞参数值影响较大,误差由未考虑温度时的38.73%降到考虑后的3.90%,而围岩参数中弹性模量误差也由7.35%降为3.90%,内摩擦角误差也由7.45%降为5.76%。并行遗传算法可以应用到隧洞参数的反演中,将计算效率提高了4倍左右。温度荷载的施加也使反演得到的计算结果更加接近实测值,反演结果更加准确。研究成果可为类似的工程提供参考。

Parallel genetic algorithm and temperature load-based study on inversion of tunnel parameter

Aiming at the problem that the temperature load is prone to be neglected in the inversion calculation made on the operation period of tunnel project and taking the deeply embedded tunnel of a large hydropower station in the Southwest China as the study case， the inversions on the parameters of the tunnel surrounding rock and the external water head during the operation of the hydropower station are made with the parallel genetic algorithm by taking the measured values of the reinforcement stress and the increments of the surrounding rock parameters as the basic data for the inversions under the consideration of the coupling effect of both the stress field and seepage field. The comparisons made between the calculation result and the measured one show that the impact from the temperature load on the tunnel parameter value obtained from the inversion is larger； from which the error is lowered from 38.73% when temperature is not considered to 3.90% under the consideration of the temperature， while the error of the elastic modulus in the surrounding rock parameters is lowered from 7.35% to 3.90% and the error of the internal friction angle is lowered from 7.45% to 5.76% as well. The parallel genetic algorithm can be applied to the inversion of tunnel parameter， by which the calculation efficiency is enhanced by 4 times. Application of temperature load can makes the calculation from the inversion more close to the measured value and the inversion result is more correct. The study result can provide a reference for the similar project.