操作参数掘进优化决策建模数据集

操作参数掘进优化决策建模数据集主要面向基于岩机模型和人工智能的数据挖掘与多元海量信息分析方法研究以及掘进装备参数优化决策研究，主要包含吉林引松供水工程TBM3施工段的掘进运行数据。原始数据的采样频率为1Hz，单个文本文件包含的操作数据最大数量每天可能高达86400个. 每个时间点运行数据有202个参数通道，包括时间戳、里程、刀盘驱动系统参数和推进系统参数等。原始数据清洗阶段，提出工作状态指标，实现工作段数据提取；制定阈值自适应规则，完成掘进循环划分；利用箱线法、移动平均法对特征数据进行异常值剔除；通过python脚本提取掘进循环上升段开始30s以及稳定段数据，通过物理定性过滤，结合方差及相关性分析，实现关冗余特征去除构建可适应模型的数据集。

Operation Parameter-Based Tunneling Optimization Decision-Making Modeling Dataset is mainly targeted at research on data mining and multi-dimensional massive information analysis methods based on rock-mechanical models and artificial intelligence, as well as research on parameter optimization and decision-making for tunneling equipment. It mainly includes tunneling operation data from the TBM3 construction section of the Jilin Yinsong Water Supply Project. The sampling frequency of the original data is 1 Hz, and the maximum number of operation data entries contained in a single text file may reach up to 86,400 per day. There are 202 parameter channels for operation data at each time point, including timestamps, mileage, cutterhead drive system parameters, propulsion system parameters, etc. In the original data cleaning stage, working state indicators were proposed to extract data from working sections; threshold adaptive rules were formulated to complete the division of tunneling cycles; outlier removal for feature data was conducted using the boxplot method and moving average method; Python scripts were used to extract the first 30 seconds of the rising segment and the stable segment data of tunneling cycles. Through physical qualitative filtering, combined with variance and correlation analysis, redundant features were removed to construct a model-adaptable dataset.

搜集汇总

背景与挑战

背景概述

该数据集包含吉林引松供水工程TBM3施工段的掘进运行数据，原始数据采样频率为1Hz，每天最多86400个数据点，涵盖202个参数通道。数据经过清洗和处理，用于掘进装备参数优化决策和人工智能研究。