An investigation on duty-cycle for particulate matter monitoring with light-scattering sensors
An investigation on duty-cycle for particulate matter monitoring with light-scattering sensors
复制标题
光散射传感器颗粒物监测占空比的研究
DOI：
10.23919/splitech52315.2021.9566363
复制
发表时间：
2021
期刊：
影响因子：
--
通讯作者：
E. Giusto
中科院分区：
文献类型：
--
作者：
P. Chiavassa;F. Gandino;E. Giusto
空气污染是我们生活的时代的一个关键现象。监控这一现象的传统方法包括使用由高成本、高精度的固定设备组成的稀疏网络。由于低端传感器的成本不断下降，现在可以实现比过去便宜得多的空气污染监测设备。尽管准确度较低，但这些设备能够正确监测颗粒物(PM)等数量。然而，在物联网(IoT)框架内运行的设备仍然存在一些问题，如电源限制、冗余信息记录、传感器老化等。为了解决这些问题，本文分析了PM传感器占空比操作的变化对聚合数据的影响，以评估信息损失。这一分析已应用于原始值和校准值。所采用的校准是以相对湿度为附加自变量的多元线性回归。结果表明，在某些情况下，即使活动时间大幅减少，也不会显著增加信息损失。采用占空比工作模式可以显著降低功耗，减缓传感器的老化，并减少冗余信息的记录和传输。此外，由于单个位置通常需要较少的样本，因此可以采用移动连续采样策略来增加PM传感器的覆盖范围。
Air pollution is a critical phenomenon of the era we live in. Traditional approaches to monitor this phenomenon involve the use of a sparse network of high-cost, high-precision fixed devices. Thanks to the decreasing cost of low-end sensors, it is now possible to implement much cheaper air pollution monitoring devices with respect to the past. Despite having lower accuracy, these devices are able to correctly monitor quantities such as Particulate Matter (PM). However, devices operating within the Internet of Things (IoT) framework still present several issues, such as power supply constraint, logging of redundant information, aging of sensors. To address these issues, this paper analyses how a change in the duty-cycle operation of PM sensors impacts aggregated data, in order to evaluate the loss of information. This analysis has been applied to both raw and calibrated values. The calibration adopted is a Multivariate Linear Regression using Relative Humidity as an additional independent variable. Results show that, in certain circumstances, even a great reduction of the active time does not significantly increase the information loss. The adoption of a duty-cycle operation mode enables a significant reduction of power consumption, slows the aging of sensors, and reduces logging and transmission of redundant information. Furthermore, since a reduced number of samples are generally required in a single location, mobile continuous sampling strategies can be adopted in order to increase the coverage of PM sensors.