进度偏差驱动的混凝土坝智能生长策略演进与差异化资源调配

Owing to the influence of resources differentiation and system uncertainty during the construction process of concrete dams, a schedule variance (SV) between the prediction before construction and the practical one is widely existed. The SV has accumulation and interlocking effects on the subsequent construction process, which will alter the growth requirements, growth characteristics and resource characteristics of subsequent dam blocks. Therefore, when SV appears, dam growth strategy should be determined in accordance with resources status and schedule requirements, at meantime, construction resources should be dynamically allocated on the premise of considering the resources differentiation..To solve the problem, some research methods (e.g. testing, data mining, intelligent computing) are employed. Starting with analyzing the construction schedule variances, the influence mechanism of their cumulative effects and chain effects on dam growth strategies is revealed, and the generation mechanism of dam growth strategies is built. Secondly, a differentiation analysis model of construction efficiency-schedule is established based on the recognition for resources differentiated properties, and the matching rules between resources differentiations and correction targets are confirmed, aimed to realize differentiated resources allocating based on correction targets and construction tasks. Meanwhile, a clustering analysis method of strategy return effects based on system status and historical experiences is proposed, which could realize the intelligent learning and dynamic evolution for dam growth strategies and resources allocation strategies. Research results can offer a novel theoretical support and solving ideas for dynamic optimization of dam growth scheme and differentiated resources allocating in complex environment on concrete dam construction site.

混凝土坝施工进程因受资源差异性和系统不确定性因素影响，与方案实施前的预测结果之间存在进度偏差。进度偏差对后续施工进程产生累积效应和连锁影响效应，改变坝体的生长需求、生长特征和资源特征。因此，发生进度偏差时，须根据系统资源状态和进度需求进行坝体生长决策并考虑资源的差异性进行动态调配。.本项目拟采用试验、数据挖掘、智能计算等方法，首先分析施工进度偏差，揭示偏差累积效应和连锁影响效应对坝体生长策略的影响机理，构建坝体生长策略产生机制；其次，建立基于资源差异化属性识别的施工效率-进度差异化分析模型，明确资源差异性与纠偏目标的匹配规则，实现基于纠偏目标与施工任务的差异化资源调配；同时，提出基于系统状态和历史经验的策略回报效应聚类分析方法，实现坝体生长策略与资源调配策略的智能学习与动态演进。研究成果可为混凝土坝施工现场复杂多变环境下坝体生长方案的动态优化及资源的差异化调配提供新的理论支撑和解决思路。

混凝土坝施工进程因受资源差异性和系统不确定性因素影响，与方案实施前的预测结果之间存在进度偏差。进度偏差对后续施工进程产生累积效应和连锁影响效应，改变坝体的生长需求、生长特征和资源特征。因此，发生进度偏差时，须根据系统资源状态和进度需求进行坝体生长决策，并考虑资源的差异性进行动态调配。.本项目结合溪洛渡、白鹤滩等高拱坝工程，通过对溪洛渡工程的后分析和跟踪白鹤滩大坝施工进程开展项目研究，首先通过构建GPS-UWB组合定位系统，进行缆机运行监测系统设计；基于采集的缆机运行海量监测数据，分析缆机工作模式与吊重时序特征，提出“孤点判别+多次迭代”方法，解决外因引起的数据波动，引入罚函数，建立吊运混凝土状态识别模型，实现对缆机实测数据的解译和效率分析；分析高拱坝施工进度风险因素和进度风险系统演化关系，建立进度风险演化仿真模型，进而建立进度偏差分析与决策模型；进行混凝土浇筑块的PBS/WBS特征分析，建立PBS/WBS节点关联关系表达，在此基础上构建浇筑块的多维特征模型；分析施工机械调配、气温、风速三种因素对实际施工效率和间隔时间的影响规律，比较不同因素对仓面施工差异性的影响；分析坝体生长路径的时空特征，提出坝块生长速度和生长策略优化方法，控制坝体生长面貌；解析多高程布设的缆机群调配与浇筑资源匹配问题结构，提出浇筑资源匹配矩阵的产生方法和资源匹配策略；建立坝体生长策略优化模型，提出基于生长策略的缆机调配优化方法；建立混凝土坝浇筑系统分析模型，提出生长策略决策目标，建立策略评价体系和策略演进模型，研发基于策略演进的浇筑仿真系统和并行计算模型，实现方案优化。研究成果可为混凝土坝施工现场复杂多变环境下坝体生长方案的动态优化及资源的差异化调配提供新的理论支撑和解决思路。