Performance Optimization of Humidity Measurement on the Surfactant-Activated ZnO–NiO Composite Humidity Sensors and Error Detection Using Machine Learning Techniques

Recently, there has been growing interest in developing highly sensitive and selective sensors for monitoring ambient relative humidity (RH). Conventional humidity sensors often face challenges, such as low sensitivity, long response times, and slow recovery. This article presents a surfactant mixed metal oxide humidity sensors that consist of nickel oxide composited with zinc oxide (ZnO) and different molar concentrations of sodium dodecyl sulfate (SDS) and cationic cetyltrimethylammonium bromide (CTAB) (SDS1:CTAB0, SDS1:CTAB0.5, SDS1:CTAB1, SDS1:CTAB1.5, and SDS1:CTAB2), namely, SC:CT-1, SC:CT-2, SC:CT-3, SC:CT-4, and SC:CT-5 to address these issues. The sensitivity of this device shows a significant improvement compared to conventional polyimide-based humidity sensors within the 20%–95% RH range. The key sensing parameters, such as response time, recovery time, and sensitivity, have been evaluated, achieving the values of 5 s, 15 s, and 49, respectively. Regression models were employed to estimate error, with the linear regression (LR) model performing the best, showing the lowest mean squared error (mse) (2.1e−24), root mean squared error (RMSE) (1.4e−12), and mean absolute error (MAE) (1.24e−12), along with a high ${R} ^{2}$ value of 0.99. Consequently, the proposed humidity sensor is a promising option for use in consumer electronics and agricultural applications.