Design and application of a high-precision counterweighted self-calibrating surface thermometer.

This article focuses on the design and evaluation of a high-precision, self-calibrating surface thermometer that employs a counterweight mechanism and platinum resistance sensors to reduce human and environmental measurement errors. The device integrates a self-weighted spring structure and a windshield with insulating materials to ensure stable thermal contact and minimize airflow-induced heat loss, achieving a maximum indication error of −0.2 °C and a repeatability standard deviation of 0.2%, outperforming conventional thermocouple-based thermometers. Software calibration using a back propagation neural network (BPNN) with a Gaussian activation function further enhances measurement accuracy by compensating for environmental factors such as wind speed. Experimental results demonstrate the thermometer’s superior stability, accuracy, and consistency, suggesting its potential as a new standard for surface temperature measurement in applications requiring high precision.