Temperature Compensation of Low-cost Sensors for Accurate Temperature Measurement

Abstract

Evapotranspiration (ET) is a critical process in agricultural system. ET states the combined water loss from the soil and plant surfaces. Predicting ET is vital for effective water resource management. Accurate temperature and humidity measurements is instrumental in calculating Et.  Temperature measurement plays a key role in various industrial, scientific, and agricultural applications. The study investigates the performance of three temperature humidity sensors. DHT11, DHT22 and LM35. The sensors are analyzed under different conditions, including compensation, and with a temperature compensation algorithm, and with proportional P and proportional-integral-derivative PID controllers. Sensors are interfaced with a Raspberry Pi 4B model. The sensors were tested in controlled environments with temperatures ranging from 10 degrees to 100 degrees C, and their accuracy was evaluated by comparing their outputs with the reference values. Results revealed that the error percentages for all sensors exceeded permissible limits when used outside their specified ranges. Incorporating temperature compensation algorithm and P and PID controllers significantly improved measurement accuracy, reducing error percentages and root mean square errors (RMSE) across all sensors. Out of the three sensors, LM35 demonstrated a 5-fold reduction in error percentage at higher temperatures compared to other methods. The improved sensor accuracy has significant implications for agricultural applications such as ET calculation, where precise temperature data is necessary. Reducing error enhances better agricultural water management strategies and overall productivity.