Application of an Intelligent Athlete State Assessment and Monitoring System Based on Multimodal Data Fusion in Evaluating the Rhythm Stability of Hurdle Runners' Inter-Hurdle Steps

Abstract

Stride frequency is critical to the competitive performance of hurdlers, and its stability directly affects athletic efficiency, energy expenditure, and injury risk. Elite athletes must precisely control their stride length (men: 1.8–2.2 m; women: 1.6–2.0 m) and stride frequency (3.8–4.2 strides/second) within a stride cycle of 0.9–1.2 seconds to avoid technical errors. Traditional assessments suffer from three major shortcomings: one-dimensional data (ignoring physiological and environmental factors), delayed feedback (post-training video analysis), and subjective bias (based on coaches’ experience).

This study adopts a three-phase framework (“Technical Development – Experimental Validation – Effectiveness Evaluation”), integrating literature review, experiments, algorithm optimization, and testing. It aims to address these issues by constructing an intelligent athlete performance assessment system through multimodal data fusion. Specific objectives: 1) Integrate biomechanical, physiological, and environmental data to establish a multidimensional rhythm stability measurement system; 2) Implement real-time multimodal data fusion and analysis, incorporating data acquisition, feature extraction, deviation detection, and feedback capabilities; 3) Validate the system’s effectiveness in supporting personalized training to enhance rhythm stability and athletic performance.