基于改进遗传交叉算子的高心墙堆石坝参数反演

来源期刊:中南大学学报(自然科学版)2016年第8期

论文作者:周伟 李少林 马刚 常晓林 胡超

文章页码:2730 - 2738

关键词:心墙堆石坝;参数反演;遗传算法;交叉算子;RBF神经网络

Key words:high central core rock-fill dam; parameter inversion; genetic algorithm; crossover operation; RBF-ANN

摘    要:受基因工程选择性克隆优秀基因片段的启发,提出一种基于基因片段差异度的自适应交叉算子(genetic crossover based on the difference of gene fragment,DGFX)。在改进的交叉算子中,随机确定基因片段长度,计算父代个体对应基因片段的差异度,根据差异度选择基因片段进行交叉操作,能有效避免近亲繁殖,减少无效交叉操作,加快收敛速度。此外,根据演化代数自适应调整基因片段长度系数,增强算法全局收敛能力。将该交叉算子与帕累托交叉算子、启发式交叉算子运用标准测试函数进行对比分析。研究结果表明:利用DGFX交叉算子时能快速收敛到全局最优解,且算法鲁棒性强、精度高。将DGFX交叉算子运用于瀑布沟心墙堆石坝堆石体力学参数反演,利用反演的力学参数进行计算,各测点计算值和实测值在发展趋势和数值上均吻合较好,说明DGFX交叉算子运用于多变量、强非线性复杂岩土工程位移反演中的优越性,具有良好的实际应用价值。

Abstract: A modified adaptive genetic crossover was provided to solve the multi-variable complex problem based on the difference of gene fragment (DGFX). In the modified crossover, the length of gene fragment was randomly determined and the difference of each gene fragment was calculated firstly. Then, the crossover point was selected according to the differences of gene fragments, which reduced the inbreeding and invalid crossover operator, and the global search capability of the algorithm was increased by the tragedy of adaptive length index. The proposed crossover was compared with two existing crossover operators (DPX and HX). A set of nonlinear test problems were used to verify the performance of the novel crossover operator. The results show that the performance of proposed crossover operator is better than or similar to those of other crossover operators and is especially effective in solving high-dimensional nonlinear problems. Then the new crossover is applied to parameters inversion of mechanical parameters of rockfill. At last, the calculated parameters are used to forecast the settlement of the monitoring points of Pubugou central core rock-fill dam. The forecasted values agree well with the measured data, which indicates that the DGFX crossover operator can be well applied to parameter inversion of complex model.

相关论文

  • 暂无!

相关知识点

  • 暂无!

有色金属在线官网  |   会议  |   在线投稿  |   购买纸书  |   科技图书馆

中南大学出版社 技术支持 版权声明   电话:0731-88830515 88830516   传真:0731-88710482   Email:administrator@cnnmol.com

互联网出版许可证:(署)网出证(京)字第342号   京ICP备17050991号-6      京公网安备11010802042557号