摘要
设计了一种基于低频(LF)唤醒技术和极限学习机(ELM)分类算法的无线定位系统。实现了低频唤醒、射频应答的电路结构和通信回路,实现了有源应答器的超低待机监听。通过比较各类型的多种射频定位算法,选用基于ELM分类的定位算法,对低频唤醒接收信号强度指示(RSSI)数据进行分类并实现定位,有效降低了定位算法在线阶段的计算量,在单片机系统中实现了实时定位计算。测试结果表明:定位系统在有效范围内定位精度可达15 cm,定位正确率可达95%以上,在定位精度和稳定性方面明显优于超高频(UHF)频段射频定位系统。
A wireless positioning system based on low-frequency（LF） wake-up and ELM classification algorithm is designed. The circuit and communication loop for LF wake-up and RF response is designed, and listening of transponder with ultra-low power consumption is achieved. By comparing numeral RF positioning algorithms, select ELM-based classification positioning algorithm, LF wake-up RSSI data is classified and located, effectively reduce amount of computations at on-line stage of algorithm, realtime positioning computation is realized on MCU systems. Test result shows that the positioning precision in effective range can reach 15 cm, correctness rate of positioning is positioning is over 95 % , it is prior to UHF RF positioning systems in positioning precision and stability.
出处
《传感器与微系统》
CSCD
2016年第3期102-104,共3页
Transducer and Microsystem Technologies
