智能船舶远程驾驶负荷累积效应实验数据及远程驾驶负荷融合评价模型

智能船舶远程驾驶员的工作负荷并非由单一因素决定，而是任务难度、操控模式和昼夜节律等多重因素动态耦合的结果。本数据集聚焦于此，系统性地探究了多因素耦合作用下的远程驾驶负荷演化规律与累积效应。数据集的核心内容为一次多因素正交实验的完整记录，该实验通过招募船员，在预设的多任务、长时程远程驾驶场景中，同步采集了覆盖整个实验过程的船舶操纵与航行轨迹数据、以及反映驾驶员状态的心电（ECG）、皮电（EDA）等多通道高频生理信号。为确保数据质量，所有数据均通过严格的实验流程控制和专业的信号处理技术进行采集与预处理。基于此数据集，本研究提取了心率变异性（HRV）等关键生理特征，并进一步构建和验证了一个能够融合多源信息、客观评估驾驶员负荷水平的机器学习模型。本数据集可为智能船舶人因可靠性、驾驶员状态监测、疲劳预警模型开发以及智能辅助系统设计等前沿领域的研究提供关键的、高质量的实证数据支撑。

The workload of remote pilots operating intelligent vessels is not dictated by a single factor, but arises from the dynamic coupling of multiple influencing factors such as task difficulty, control mode, and circadian rhythm. This dataset centers on this topic, systematically investigating the evolutionary patterns and cumulative effects of remote piloting workload under the coupling of multiple factors. The core of this dataset consists of the complete records of a multi-factor orthogonal experiment. In this experiment, seafarers were recruited to participate in a preset multi-task, long-duration remote piloting scenario, during which synchronously collected data included ship maneuvering and navigation trajectory data spanning the entire experimental process, as well as multi-channel high-frequency physiological signals reflecting the pilot’s status, such as electrocardiogram (ECG) and electrodermal activity (EDA). To ensure data quality, all data were collected and preprocessed through strict experimental process control and professional signal processing techniques. Based on this dataset, this study extracted key physiological features such as heart rate variability (HRV), and further constructed and validated a machine learning model that can fuse multi-source information to objectively evaluate the pilot’s workload level. This dataset can provide critical, high-quality empirical data support for research in frontier domains including human factors reliability of intelligent vessels, pilot state monitoring, fatigue warning model development, and intelligent auxiliary system design.

搜集汇总

背景与挑战

背景概述

该数据集聚焦智能船舶远程驾驶员的负荷累积效应，通过多因素正交实验采集了船舶操纵、航行轨迹以及心电、皮电等多通道生理信号，用于研究多因素耦合下的负荷演化规律。基于这些数据，研究构建并验证了一个融合多源信息的机器学习模型，以客观评估驾驶员负荷水平，为智能船舶人因可靠性、疲劳预警等领域提供高质量实证支撑。数据集由武汉理工大学智能交通系统研究中心创建，数据量28.68GB，包含431个文件，源自国家重点研发计划项目。