Agricultural Data Rectification Algorithm Based on Internet of Things

Abstract

Agricultural information data in Internet of Things is the basis of abnormal environment warning system, however, these data in the measurement process will inevitably contain random error and gross error. For the problem of exceptional data existing in wireless sensor networks, a data rectification model based on the detection of gross error is proposed, which can automatically eliminate and compensate the gross error in measured data, in order to achieve the purpose of accurate data rectification. First, utilize the statistical tests method to detect and identify gross error in measured data , and the method can detect more than three gross errors, which show superiority in the number of gross error than other gross error detection methods. Then, median filter is exploited to estimate the value of gross error identified. At last, three order exponential smoothing is used to data reconciliation. Experimental results show that this method can achieve effective performance.