高心墙堆石坝心墙砾石土智能压实理论与方法研究

This project aims at the main problems of scientific research of core gravelly soil intelligent compaction, based on the actual high rockfill dam construction, researching on the theory and method of core gravelly soil intelligent compaction in high rockfill dam. This project studies real time monitoring method of core gravelly soil blending process , including tile vertical mining construction parameters real-time detection and roller blending results real-time detection, to improve the sensing system of core gravelly soil construction process ; This project also studies multi-source construction information integration method of core gravelly soil, to realize the timely, continuous and complete integration of multi-source construction information of core gravelly soil , on the dimension of time, space and attribute; It uses intelligent method to do deep excavation, finding knowledge of Interaction between construction information and compaction quality, to realize the real time evaluation of compaction quality based on real time information and the compaction quality prediction according to the regular pattern of construction information change and development; It studies the intelligent decision making method for optimal design of construction scheme and dynamic adjustment of construction process, based on real time compaction quality evaluation and prediction, to realize the effective and timely feedback control for the compaction construction of core gravelly soil; It builds the system of perception-integration- evaluation and prediction- intelligent decision -feedback control, to realize intelligent compaction. The development of this project helps to solve the key scientific problems in the intelligent filling and compaction of the core gravelly soil, enriching and developing the existing intelligent dam technology system, and improves the construction and management level of earth rock dam in China. It is of great theoretical and practical significance in the world.

本项目拟针对心墙砾石土智能填筑压实研究中存在的关键科学问题，开展高心墙堆石坝心墙砾石土智能填筑压实理论与方法的研究。研究心墙砾石土掺配工艺实时监测方法，完善心墙砾石土施工过程感知方法体系；研究心墙砾石土多源施工信息集成理论方法，实现在时间维度、空间维度和属性维度上及时、连续、完整的集成多源施工信息；采用智能方法进行深度挖掘，发掘反映施工信息同压实质量之间相互作用关系的知识，实现结合实时获取的真实施工信息对压实质量进行实时评估、根据施工信息的发展变化规律对压实质量进行预测；研究基于压实质量评估和预测的施工方案优化设计和施工过程动态调整的智能决策方法，实现对于心墙砾石土填筑压实施工的有效、及时的反馈控制；构建施工感知-信息集成-评估与预测-智能决策-反馈控制的体系，以实现智能压实。本研究项目的开展，丰富和发展了现有的智慧大坝技术体系，在国内外具有创新性，具有重大的理论意义和现实意义。

本项目针对心墙砾石土智能填筑压实研究中存在的关键科学问题，开展高心墙堆石坝心墙砾石土智能填筑压实理论与方法的研究。本项目的研究内容为开展心墙砾石土施工感知方法的研究、心墙砾石土多源施工信息集成理论方法的研究、多源施工信息深度挖掘理论方法的研究、以智能挖掘为基础开展压实质量评估与预测理论方法的研究、应用智能技术进行心墙砾石土施工决策的理论方法研究、结合实际高砾石土心墙堆石坝工程进行工程应用分析，从而改进心墙砾石土智能压实理论方法。取得了如下的总体研究果：进行基于不规则颗粒离散元的砾石土三轴数值模拟和基于施工全过程参数的堆石坝掺砾土心墙压实质量实时评价研究，为改善心墙砾石土力学性质和碾压堆石坝施工技术提供参考。基于有限元法的心墙砾石土压实特性分析，为控制心墙砾石土的压实质量提供科学的理论依据。进行复杂施工条件下无人碾压机群协同全覆盖路径规划研究，对于保证无人碾压机在大坝复杂施工条件下的碾压质量和进度具有重要意义。提出基于ACGWO-SVR的高寒地区心墙堆石坝压实质量评价模型，通过稳健优化确定粗集料轧制参数控制标准，对于降低现场数据对噪声的敏感性，提高压实质量的测量精度和稳定性具有重要意义；进行了基于RF-ACGWO的压实质量实时评价具有较高的鲁棒性和泛化能力的研究，在RF-ACGWO中，RF预测压实质量，而ACGWO提高了传统RF参数选择的效率和准确性，提高了模型的泛化性和鲁棒性。提出了基于增强现实的心墙堆石坝碾压施工过程可视化方法，可实现增强现实环境下心墙堆石坝碾压施工过程的可视化展示与交互分析。提出了实时监控下基于二进制蝴蝶优化算法与特征选择的堆石坝压实质量评估方法，具有更佳的评估性能。上述研究成果行成了成套的心墙砾石土智能填筑压实理论与方法，具有重大的理论意义和现实意义。