基于改进卡尔曼算法的室内温度数据融合

Indoor Temperature Data Fusion Based On Improved Kalman Algorithm

摘要:

针对传统建筑物内部空间结构复杂，布线成本高，数据采集精度不高等问题，采用LoRa无线通信技术构建传感器网络，主要用于监测室内温度参数的变化;对于传统的卡尔曼数据融合结果存在较小波动的现象，引入了孤立森林算法，提出了基于改进的卡尔曼滤波算法的室内温度数据融合算法；通过在采集到的数据集中随机添加扰动样本和畸变数据，对三种算法产生的误差进行比较，改进的卡尔曼数据融合算法在有扰动样本的情况下，误差范围控制在-0.12-0.1之间，在带有畸变数据时，误差范围在-0.03至0.14之间，均远小于传统的卡尔曼数据融合算法和平均值算法；实验仿真的结果表明，改进的算法提高了室内温度数据采集的鲁棒性和准确性。

Abstract:

For the problems of complex internal space structure of traditional buildings, high wiring cost and low data acquisition accuracy, LoRa wireless communication technology is used to build a sensor network, which is mainly used to monitor the changes of indoor temperature parameters; for the phenomenon of small fluctuations in the traditional Kalman data fusion results, the isolated forest algorithm is introduced, and the indoor temperature data fusion algorithm based on the improved Kalman filter algorithm is proposed The error range of the improved Kalman data fusion algorithm is controlled between -0.12 and 0.1 with perturbed samples and -0.03 to 0.14 with distorted data, which are much smaller than the traditional Kalman data fusion algorithm and the mean The results of experimental simulation show that the improved algorithm improves the robustness and accuracy of indoor temperature data acquisition.