Ground-based air temperature measurements are highly susceptible to solar and environmental long-wave radiation, often leading to temperature deviations of up to approximately 1 °C. This study proposes a low-uncertainty air temperature measurement instrument specifically designed to minimize such radiation-induced temperature deviations. Computational fluid dynamics (CFD) simulations were first performed to optimize the structural design of the instrument, thereby enhancing internal airflow and improving both radiation shielding efficiency and convective heat dissipation. Subsequently, the radiation-induced temperature deviations of the optimized structure were quantitatively analyzed under various environmental conditions. A multi-layer perceptron (MLP) neural network was then employed to develop a temperature deviation correction model using the CFD-generated dataset. Finally, field comparative experiments were conducted using a 076B aspirated temperature measurement instrument as a reference. Experimental results show that, before correction, the proposed instrument exhibited a maximum radiation-induced temperature deviation of 0.43 °C and a mean temperature deviation of 0.31 °C. The root mean square error (RMSE), mean absolute error (MAE), and correlation coefficient (r) between the experimental and predicted radiation-induced temperature deviations were 0.08 °C, 0.07 °C, and 0.99, respectively. After applying the correction model, the maximum radiation-induced temperature deviation decreased to 0.12 °C, and the mean deviation decreased to 0.02 °C, demonstrating excellent consistency. In conclusion, the proposed instrument achieves efficient ventilation and low-uncertainty temperature measurement without relying on mechanical components.
扫码关注我们
求助内容:
应助结果提醒方式:
