基于双波段近红外光谱的森林凋落物含水率测量方法

Measurement Method for Moisture Content in Forest Litter Based on Dual-band Near-infrared Spectroscopy

摘要:

在森林防火监测中,地表凋落物含水率是评估区域火险等级的关键指标。现有基于单波段近红外光源的含水率测量系统易受环境光干扰,导致自然光照变化条件下测量误差较大。本研究通过近红外光谱分析实验发现,凋落物中水分子对中心波长1450nm波段近红外光的吸收强度显著强于中心波长1350nm波段。基于此原理,本研究设计了一种基于双波段近红外法的凋落物含水率测量系统。与单波段系统不同,该系统并非依赖单一波段的模数转换器(ADC)数值进行含水率标定,而是利用中心波长1450nm与1350nm波段ADC数值的差值进行标定。该系统硬件由单片机、双波段红外发射器、ADC采集模块、串口通信模块、外置电源及不锈钢防护外壳等组成。在常温及室内稳定光照条件下标定实验表明,系统ADC采集值与凋落物含水率之间关系符合三次多项式模型(R2=0.9859)。为验证双波段设计在动态自然光照下的测量鲁棒性,将充分吸水的凋落物样本置于户外自然干燥,定时检测其含水率与质量变化,共获取23组样本数据。通过对比中心波长1350nm与1450nm波段ADC数值变化曲线,证明双波段系统在自然光干扰下仍能保持较高测量精度。实验结果表明,该传感器的平均绝对误差为2.74%。本研究所设计的双波段近红外测量系统有效克服了单波段系统在环境光变化下的误差问题,显著提升了抗干扰能力。

Abstract:

In forest fire monitoring, the moisture content of surface litter is a key indicator for assessing regional fire danger levels. Existing moisture content measurement systems based on a single near-infrared (NIR) light band are susceptible to ambient light interference, leading to significant measurement errors under natural light variations. Through near-infrared spectroscopy experiments, it was found that water molecules in litter exhibited significantly stronger absorption intensity at 1450nm band compared with 1350nm band. Leveraging this principle, a dual-band NIR measurement system for litter moisture content was designed. Unlike single-band systems which relied solely on the analog-to-digital converter (ADC) value of one wavelength for calibration, this system utilized the difference between the ADC values at 1450 nm and 1350nm. The hardware comprised a microcontroller, dual-band infrared emitters, an ADC acquisition module, a UART communication module, an external power supply, and a stainless-steel protective enclosure. Calibration experiments under stable room temperature and indoor lighting showed that the relationship between the system's ADC values and litter moisture content fitted a cubic polynomial model (R2=0.9859). To verify the system's robustness against dynamic natural light during measurement, water-saturated litter samples were dried naturally outdoors while their moisture content and weight were monitored periodically, yielding 23 datasets. Comparative analysis of ADC value trended at 1350nm and 1450nm demonstrated that the dual-band system maintained high accuracy despite natural light interference. Experimental results indicated a mean absolute error ( MAE) of 2.74% . The designed dual-band NIR measurement system effectively overcame the limitations of singleband systems under ambient light variations and significantly enhanced interference resistance.