基于遗传算法的软土地基沉降预测

Prediction of soft soil foundations settlements based on genetic algorithm

基于自然进化原理的遗传算法是一种解决岩土工程复杂非线性问题十分有前景的方法。对遗传算法中一些具有代表性的算法作了简要分析和评论。软土地基沉降受众多因素的影响,在时序上表现出复杂的非线性特征。针对一个加固纠偏工程的沉降预测问题,运用MATLAB编写遗传算法程序进行了预测分析。计算结果与实测资料基本一致,它在全局优化、非线性优化、多参数优化等方面表现出了传统算法无法比拟的优势。

Genetic algorithm based on natural evolutionism is a good method to resolve complicated geotechnic nonlinear problems. Some classical works on genetic algorithm are introduced and remarked. The settlement of a soft soil foundation often manifests as complicated features at time sequence due to the influence of many factors. In the settlement prediction of a reinforcement and slant rectification, genetic algorithm programme coded by MATLAB predicts the settlement. Calculation result is consistent with the result from actual measurement. It indicates that genetic algorithm has much more superiority than traditional algorithm in global optimization, nonlinear optimization, multiparameter optimization, and so on.