随着传感与检测技术的迅速发展，实测数据分析在复杂问题实验研究或大工程系统研究中起着日益重要的作用。本项目拟进行海量工程实测数据反演识别新方法的研究，将量纲分析与数据挖掘技术相结合，对海量数据中主干规律与细致特征进行分层次提取，为具有非线性、多参量、时变性特点的复杂工程系统的实验数据分析提供一种新思路。本项目将这一方法用于盾构装备载荷与能效的力学分析研究，掘进载荷是贯穿装备运行始终的核心力学量，也是装备能效控制与安全状态监控的基础参量。但是影响载荷的因素多，且工况复杂，目前行业还主要参照Krause经验公式，估值过于宽泛。建立合理有效的载荷模型是该领域需要解决的难点问题。本项目拟抓住装备与地质间相互耦合作用的力学本质，揭示地质条件、装备操作状态、装备结构特征这三类核心要素与载荷间的相互影响规律，建立装备载荷与掘进能效的力学模型，提出关键控制量的优化配比，为装备设计与安全施工提供科学依据。

With the rapid development of sensing and detection techniques, measured data analysis plays an increasingly important role in the research of complex experiments and heavy engineering equipments. The project intends to work out a new method for massive measured data inverse identification, in which dimensional analysis and data mining techniques will be combined. Trunk laws and detailed characteristics of massive data can be hierarchically extracted. The method will provide a new way for experimental data analysis of complex engineering systems with non-linear, multi-parameters, time-varying characteristics. The method will be used into the mechanical analysis of shield tunneling loads and energy efficiency. The loads are not only core mechanics parameters working throughout equipment running process, but also basic parameters in effective control and security status monitoring. However, there are lots of factors which can affect loads and the real tunneling conditions are very complex. As the main reference in industry, Krause empirical formula gives a broad estimated range. To establish a reasonable and effective load model is the aporia in tunneling field. The project intends to seize the mechanics essence of mutual coupling between equipment and geology and to reveal the influence of three core elements on loads including geological conditions, equipment operating status and structural features. Then a mechanical model of loads and energy efficiency will be established to give a optimum ratio among the critical control parameters. The results will provide a scientific basis on equipment design and construction safety.