摘要
施工进度仿真是高拱坝施工进度控制的重要手段。当前高拱坝施工进度仿真分析中,仿真施工参数基本都是根据施工经验参考近似工程选取,然而不同工程之间施工参数差距较大,可能导致仿真计算结果的失真。针对以上问题,本文提出基于贝叶斯更新理论,根据现场实际施工信息,实时更新仿真施工参数,以进行仿真计算分析。然而贝叶斯更新过程需要大量精确的统计数据,在实际工程中,由于人为及机器等多种因素,收集到的数据往往不够充分且不够精确。为此,本文将改进重抽样(Bootstrap)方法引入到贝叶斯更新中,提高了样本数据的代表性,使得样本统计参数的估计值更为准确。仿真分析结果表明,相比于传统方法,采用本文提出的施工参数实时更新方法,可以提高仿真参数精度。
Construction simulation is an important approach for high arch dam construction schedule manage- ment. For the construction simulation for high arch dams in current days, the construction simulation param- eters are determined based on construction experience according to approximate projects. But the construc- tion parameters vary between different projects, which may lead to the distortion of construction simulation. In the light of this problem, this research proposes the method that, the Bayesian updating algorithm is uti- lized to update simulation parameters based on the actual construction information. The Bayesian updating process needs a large number of accurate statistical data, but the collected data is not enough and accu- rate, because of human and mechanical factors. Therefore, the modified Bootstrap is introduced in Bayesian updating to improve the representative of simple data, and the estimated value of simple statistical data is more accurate. The simulation result shows that, compared with traditional methods, the real-time updating method of construction simulation proposed in this paper could improve the accuracy of simulation parame- ters.
出处
《水利学报》
EI
CSCD
北大核心
2016年第4期473-482,共10页
Journal of Hydraulic Engineering
基金
国家自然科学基金重点项目(51339003)
国家自然科学基金创新研究群体科学基金项目(51321065)
国家自然科学基金青年科学基金项目(51409186)
