Dynamic compensation and correction of measurement error in soil moisture sensor based on FLANN algorithm

To address issues such as nonlinear error and time-varying drift caused by susceptibility of soil moisture sensor to complex soil environments and their own dynamic characteristics, a dynamic compensation and correction method for soil moisture sensor measurement error based on FLANN algorithm has been proposed．A soil moisture sensor was designed based on standing wave ratio principle, comprising a signal source, transmission line, and probe．Using FLANN algorithm, an error compensation model was constructed with original measurement values of soil moisture sensor and environmental interference factors as inputs, and true values as outputs．Static interference error caused by environmental factors has been compensated through iterative optimization of weight parameters．To address dynamic response lag errors of soil moisture sensors, a mathematical model of an inverse system compensator has been constructed, integrating FLANN algorithm with backpropagation（BP）neural network to iteratively approximat optimal weight values, thereby achieving dynamic compensation and correction．Test results have shown that this method significantly improved measurement accuracy, with a normalized mean square error of ＜−21 dB after compensation and correction, enabling adaptation to complex soil environments and meeting long-term operational requirements．