摘要
采用常规监理旁站方式监控土石坝碾压参数以及事后试坑取样检测的质量管理手段,往往有较大的人工误差。依托土石坝施工质量实时监控技术,获得了施工仓面任意位置处的碾压参数值(碾压速度、激振力、碾压遍数和压实厚度)。在分析碾压参数与压实质量标准(干密度或压实度)之间相关性的基础上,采用三层前向式ANN模型,建立碾压参数和压实质量标准之间的非线性映射关系,并提出了全仓面碾压质量评估的具体流程,获得仓面任意位置处的压实度以及全仓面的碾压质量达标率。该方法可有效避免以有限个试坑检测样本评价整体仓面碾压质量的片面性,为现场施工质量控制提供了新的途径。
Site supervision of roller compacting parameters and pits sampling are conventional ways for the quality control of earth dam construction.However,personal errors and slips of quality management may often appear.By using real-time quality monitoring technology of earth dam constructiont,he real-time compaction parameters at any position of the working area including rolling speed,vibration force,rolling passes and compacted thickness can be collected.Based on the correlation analysis between compaction parameters and quality index（dry density or compaction degree）,the nonlinear mapping relationship between input（compaction parameters） and output（quality index） was established using an three layers feed-forward artificial neural networks model,and a process of compaction quality evaluation of the entire rolled working unit was suggested to estimate the compaction degree at any position and the proportion of the acceptable area.This method can effectively avoid the incorret assessment for the rolling quality of the entire working unit according to limited observations of pit sampling.
出处
《水利学报》
EI
CSCD
北大核心
2010年第6期720-726,共7页
Journal of Hydraulic Engineering
基金
国家自然科学基金青年基金项目(50709024)
教育部新世纪优秀人才计划资助(NCET-08-0391)
关键词
土石坝施工
碾压
质量控制
质量评估
实时监控
神经网络模型
Earth dam construction
compaction
quality control
quality evaluation
real-time monitoring
artificial neural networks
