Human Vital Signs Detection Based on E-WOA Algorithm

This paper presents an enhanced method for non-contact detection of human vital signs using FMCW radar. A key challenge in this field is accurately separating weak respiratory and heartbeat signals from environmental noise and body movement artifacts. To address this, the authors propose a novel algorithm combining Variational Mode Decomposition with an Enhanced Whale Optimization Algorithm. The proposed E-WOA is designed to adaptively optimize VMD’s crucial parameters, overcoming the defects of the standard WOA, such as slow convergence and susceptibility to local optima. This is achieved by introducing a new population initialization, a pooling mechanism, and improved search strategies (migration, priority selection, and enriched encirclement), which enhance the algorithm’s global exploration ability. The performance of E-WOA was validated against the original WOA on benchmark functions, demonstrating superior convergence and accuracy. Subsequently, real-world experiments using a millimeter-wave radar confirmed that the E-WOA-VMD method significantly outperforms the traditional WOA-VMD approach. It achieves lower error rates and greater stability in estimating respiration and heart rates, proving its effectiveness and potential for reliable, non-contact health monitoring.