XU Jiang-chun, YU Li-ling, YUE Qiu-yan, et al. PM2.5 concentration prediction system based on improved Kalman filtering[J]. 2017, 36(7): 106-109. DOI: 10.13873/j.1000-9787(2017)07-0106-04.
改进型卡尔曼滤波的PM2.5浓度预测系统
摘要
针对目前室内空气净化器功能单一、自动调节能力弱的特性
设计了一种改进型卡尔曼滤波对PM 2.5浓度进行预测的系统
为净化器提供可靠的调节参数。选用S3C2440芯片
通过外设采集PM2.5浓度和温度
以及对历史数据进行分析。通过改进型卡尔曼滤波对历史数据、PM2.5浓度以及温度进行融合
可以准确得到下一时刻的浓度值。测试结果表明:系统运行稳定
算法跟踪效果明显
鲁棒性强
可以准确预测出下一时刻的PM2.5浓度
并且预测误差在2.0%以内
满足设计要求。
Abstract
In view of the characteristics of current indoor air purifier function is single and the automatic regulation ability is weak
an improved Kalman filtering system is designed to predict the concentration of PM2. 5
which provides a reliable regulation parameter for air purifier. The S3C2440 chip is used to collect PM2. 5concentration and temperature
and analyze on historical data. On the basis of the system
through the improved Kalman filter to fuse historical data
PM2. 5 concentration and temperature can be accurately obtained concentration at the next moment. The results of test show that the system runs stably with its effective algorithm
strong robustness
and it can accurately predict the PM2. 5 concentration at the next time point. The prediction error is less than 2. 0 %