基于毫米波雷达的呼吸检测与分类算法研究

Research on Breath Test&Classification Algorithm Based on Millimeter-wave Radar

与传统的呼吸检测技术相比,毫米波雷达检测在用户体验、检测效率与安全性方面表现更优,未来前景十分广阔。利用77 GHz毫米波雷达发射并处理混频信号,再采用快速傅里叶变换以及相位提取与解卷绕算法,经过IIR滤波器提取呼吸信号,最终对不同的呼吸状态进行分类,实现了从检测到呼吸状态分类的完整过程。结果表明,呼吸检测方法有很强的稳定性,并且支持向量机算法的分类准确率为95%,可以有效识别不同的呼吸状态。

Compared with the traditional breath detection technology,the millimeter-wave radar detection method has better performance in terms of user experience,detection efficiency and safety,and its future prospects are very broad.The 77 GHz millimeter-wave radar was used to transmit and process the mixed signal,and then the fast Fourier transform and phase extraction and unwrapping algorithm were used to extract the breathing signal through the IIR filter.The classification was carried out,and the complete process from detection to classification of breathing state was realized.The results show that the breathing detection method has strong stability,and the classification accuracy of the support vector machine algorithm is 95%,which can effectively identify different breathing states.