A Dehumidification Efficiency Prediction Model Based on Polynomial Regression for Desiccant Wheels

Desiccant wheels are important devices in energy-saving air-conditioning systems and it is important to know the dehumidification efficiency of desiccant wheels for their design and optimization. It is faster and more electricity-saving to obtain the efficiency through a prediction model compared with the experimental methods. In this work, a dehumidification efficiency prediction model based on polynomial regression for desiccant wheels is established. The model is aimed at predicting the dehumidification efficiency at different working conditions through a limited set of experimental data. As tested by experimental data, the absolute and the relative prediction errors of the model after modification for most data points are within $\pm 2\%$ and $\pm 7\%$. In addition, the test results show that adding cubic terms of some variables to the quadratic polynomial regression model helps to improve the prediction accuracy, which shows the improvement potential of the model. In conclusion, the presented model can provide a reference for the performance prediction and optimization of the wheel.