基于RBF神经网络算法的被动红外探测器设计及单片机实现

RBF Neural Network Arithmetic of Passive Infrared Detector and Realization by MCU

传统被动红外探测器容易受到环境温度,环境光线强度影响,误报率高。对此提出采用RBF神经网络算法来融合环境温度,以光线强度参数来改善对传感器的非线性补偿效果方法,并提出单片机实现方式。实验数据表明该方法能够显著提高探测器的环境适应能力,具有很强的实用性。

The origin passive infrared detectors is easily jammed by factors such as environmental temperature and environmental light intensity.This paper provides a method of improving the performance of passive infrared detectors by data fusion and non-linearity compensation which is based on Radial Basis Function(RBF) neural network arithmetic.The realization by MCU is presented too.The result of experiment data shows that this way can markedly improve the detector's performance and is proved to be workable.