【摘要】 结合遗传进化算法的全局优化思想和时间序列分析的基本理论,提出了一种新的基于遗传算法的建筑物沉降非线性动力学行为预测模型进化识别算法。设计了模型结构和参数分别进化,共同识别方案,实现对非线性时间序列分析模型结构和参数进行全局最优搜索。实例分析结果表明:该方法具有较好的预测精度和推广预测能力,为高大建筑物沉降预测提供了一个有效的分析工具。更多还原

【Abstract】 This paper presents a new evolutionary method,combining the global optimal techniques of genetic algorithm and the elementary theory of time series analysis,to model the nonlinear dynamic behavior of building settlement.A new stepped evolutionary scheme is designed to search the global optimal structure and parameters of the nonlinear time series model.Application to settlement prediction of a high building is also given and the satisfying results show the great performance of the new method.It provides an efficient analysis tool for the prediction of settlement of high buildings.更多还原