为了满足自动监测游泳状态应用需求,设计了一种无线溺水特征识别系统.该系统能在不影响正常游泳的情况下通过自动化手段实时、准确监测游泳者的游泳状态,并对异常情况进行报警.系统的硬件部分由信号采集与预处理模块、无线传输模块和主控模块等组成.主算法部分分为串口数据接收模块、数据预处理模块、溺水特征识别模块和控制模块等.其中溺水特征识别模块是基于人工免疫克隆选择算法进行实现的,与其他识别算法相比其具有天然的模式识别特性,并能提取出常规算法不易提取的特征.由实验结果可看出,该系统可准确、迅速地识别出溺水和非溺水特征.
A wireless drowning features recognition system is designed to meet the application requirements of automatic monitoring swimming states. The system is not only capable of real-time monitoring of the status of swimmers accurately and automated without affecting normal swimming, but also of abnormal alarms. The system hardware consists of signal acquisition and pre-processing module, a wireless transmission module and control module. The main part of the algorithm contains serial data receiver module, data preprocessing module, drowning feature recognition module and the control module. Drowning feature recognition module is based on the artificial immune clonal selection algorithm, which has a natural recognition features pattern that can easily extract features that are difficult to extract by other recognition algorithms. Experiments show that the system can accurately and quickly identify the characteristics of non-drowning and drowning.