Air Temperature Error Correction Based On Solar Radiation In An Economical Meteorological Wireless Sensor Network

With the fast development of wireless sensor network (WSN), a novel method named the meteorology wireless sensor network has been proposed for the purpose of meteorology factors monitoring. It will reduce the cost of the project by using inexpensive temperature sensor for the sensing of air temperature (AT). Nevertheless, the cheap sensor on the sensing node can be easily influenced by environmental factors, such as solar radiation (SR), precipitation, wind, etc. The data sensed by the sensing node were inaccurate. Therefore, this paper proposed a method to decrease the error of sensed AT, taking SR into consideration. The main achievements in this paper are as follow:(1) A correction method of time series is presented. Firstly, we use spline interpolation to unify the time sample of all data. Then, knowledges of atmospheric physics are used to analysis the relationship between AT sensed by node (NodeAT) and normal AT collected by automatic weather station (AwsAT). Lastly, we correct the time difference between NodeAT and AwsAT. The consequence of experiment shows that the correction of time difference can decrease the deviation between NodeAT and AwsAT at night.(2) The correction method is established based on the correction of time deviation by taking SR into consideration. Firstly, we calculate the deviation between normal AT and time shifted NodeAT, and find the relationship between AT error (ATE) and SR. Then we analysis the data and get a table contains the relationship of ATE and SR. Finally, this table is utilized to correct ATE value of NodeAT. The experiment result shows that error of NodeAT is corrected effectively.(3) A formula used to calculate corrected AT is built based on the table of SR-ATE.In this part, we first get an expression to represent the relation of ATE and SR by means of curve fitting. Then the formula used to calculate corrected AT is deduced according to the correcting method. Lastly, we have some comparison between the correcting result based on this formula and result based on discrete table of SR and ATE, and find that the estimation formula not only can cut down the temperature deviation of NodeAT, but also can improve the accuracy of corrected temperature further.