多源传感器监测及数据融合方法研究

Research on multi-source sensor monitoring and data fusion methods

针对多源传感器系统中存在着异常数据处理不及时、数据融合计算效率较低等问题,本文提出了一种基于LSTM和扩展卡尔曼滤波的多源传感器监测及数据融合方法,并以空气质量监测为例详细论述了该研究的应用。通过物联网采集模块,实现了传感器数据测量和传输,选用LSTM神经网络来对其进行异常数据处理,通过扩展卡尔曼滤波算法对测量过程产生的动态误差进行补偿,进而实现传感器目标状态高精度的测量,达到有效去除噪声影响的目的。最后,经过实验结果分析,表明该方法有较好的实践应用效果。

Aiming at the problems of untimely abnormal data processing and low efficiency of data fusion calculation in multi-source sensor systems,a multi-source sensor monitoring and data fusion method based on LSTM and extended Kalman filtering is proposed,and the application of this research is discussed in detail using air quality monitoring as an example.Through the IoT acquisition module,the sensor data measurement and transmission are realized,and the LSTM neural network is selected to process the abnormal data.The extended Kalman filter algorithm is used to compensate the dynamic error generated by the measurement process,which can achieve high-precision measurement of the target state of the sensor and achieve the purpose of effectively removing the influence of noise.Experimental results analysis shows that the proposed method has good practical application effects.