Experimental investigation of pedestrian locomotion: Gait characteristics and walking speed under the combined effects of roll and pitch motions

To obtain the walking speed and gait characteristics of pedestrians in a ship's motion environment, this study conducted a series of experimental investigations. Experiments were designed with single-angle roll/pitch and combined rolling and pitching conditions. Based on the extracted trajectories, the walking speed distribution of pedestrians under different conditions was obtained. An automatic extraction method was used to measure gait parameters such as step length, step time, and step width. A linear fitting approach was applied to study the correlation between walking speed and step length, step time, and step width. Furthermore, a pedestrian walking speed regression prediction model was developed based on Support Vector Machines (SVM). The model was trained using data obtained from the experiments, and the validation results showed that the model could effectively predict pedestrian walking speed in a ship's motion environment. Based on the SVM model, the critical angles at which pedestrians might lose their ability to move were analyzed. This study provides data support for improving ship motion capability assessment and for expanding, supporting, and validating existing pedestrian evacuation models. It also contributes to enhancing the safety management level of ship pedestrian.