无线传感网络中多目标识别与定位系统设计

Wireless Sensor Networks Multi-target Detection and Location System Design

设计了一种基于无线传感网络的多目标识别与定位系统。系统采用DSP控制器作为传感节点的核心,实时采集移动目标的红外信号、磁信号和声音信号,并对所采集的目标信号进行预处理,通过无线传感器网络将所采集的目标信息上传至上位机,通过上位机调用基于Matlab的神经网络目标识别算法和定位算法分别实现对移动目标的识别与定位。分别进行了室内单人定位和小车定位的多目标探测实验,结果表明,设计的基于无线传感网络的多目标识别定位系统对目标的识别正确率均达到92%以上,定位效果的均方误差小于0.22 m,达到了探测系统的技术指标要求。

A new system design to identify and locate multiple objectives in WSNs（Wireless Sensor Networks） was proposed. This system adopted DSP controller as the center of sensor nodes and collected infrared, magnetic and acoustic signals of moving targets for preprocessing. The signals were transmitted to upper computer and then the targets were recognized and located utilizing objective identification and location algorithms of neural network via Matlab. Ultimately, experiments were executed to verify the system effectiveness in terms of single objective identification, multi-objective location and moving target location. The results indicated that the system proposed in this paper can achieve an accuracy rate of above 92% in objective identification and a mean squared deviation of below 0.22 m in location. Therefore, the detection system satisfied the specifications.