Research On Influence Factors And Compensation Methods Of Soil Moisture Sensor

Plants grow in the soil and they need the soil moisture to irrigate constantly.It is significantly to study the condition of the soil moisture to the growth of the plants.The soil moisture sensor can test the content of soil moisture in real time,so more and more soil moisture sensors are used in the farming areas.There are many types of soil moisture sensors,FDR soil moisture sensor is easy for use,and has the advantages of good real time performance,so it has been widely used in soil moisture detection.But according to the study of the predecessors and the reaction of the users found that FDR sensor is easily influenced by the interference factors such as the temperature and the hardness of the soil.It will affect the final measurement results,if not to compensate the interference factors,and as a result,it will influence the judgment of the content of the soil moisture.After analysising and researching on the influence factors,the paper focuses on the influence of temperature on the FDR soil moisture sensor.Using two element regression analysis and BP neural network for temperature compensation of FDR soil moisture sensor,and simulated by using Matlab software.Temperature sensitivity coefficient is 3.9x10-3/ ℃before compensation,after temperature compensating by using two element regression analysis method, the temperature sensitivity coefficient decreased to 1.3 x 10-3/ ℃; after temperature compensating after by using BP neural network method,the temperature sensitivity coefficient decreased to 5.85x10-3/ ℃.So after temperature compensation by the two methods, temperature sensitivity coefficient of FDR soil moisture sensor reduces a little,and the sensor’s temperature stability has a certain degree of improvement. In this paper comparison were made between the effect of compensation of two element regression analysis method and the effect of compensation of BP neural network method. The results show that, the average error compensation of the two element regression analysis method is0.1224%,the BP neural network compensation method is 0.0548%,and the error compensation is one order of magnitude smaller than two element regression analysis method, and the error compensation of BP neural network distribute more uniformly thanthe error compensation of two element regression analysis method.