Resistive Sensing of Seed Cotton Moisture Regain Based on Pressure Compensation

The measurement of seed cotton moisture regain (MR) during harvesting operations is an open and challenging problem. In this study, a new method for resistive sensing of seed cotton MR measurement based on pressure compensation is proposed. First, an experimental platform was designed. After that, the change of cotton bale parameters during the cotton picker packaging process was simulated through the experimental platform, and the correlations among the compression volume, compression density, contact pressure, and conductivity of seed cotton were analyzed. Then, support vector regression (SVR), random forest (RF), and a backpropagation neural network (BPNN) were employed to build seed cotton MR prediction models. Finally, the performance of the method was evaluated through the experimental platform test. The results showed that there was a weak correlation between contact pressure and compression volume, while there was a significant correlation (p < 0.01) between contact pressure and compression density. Moreover, the nonlinear mathematical models exhibited better fitting performance than the linear mathematical models in describing the relationships among compression density, contact pressure, and conductivity. The comparative analysis results of the three MR prediction models showed that the BPNN algorithm had the highest prediction accuracy, with a coefficient of determination (R-2) of 0.986 and a root mean square error (RMSE) of 0.204%. The mean RMSE and mean coefficient of variation (CV) of the performance evaluation test results were 0.20% and 2.22%, respectively. Therefore, the method proposed in this study is reliable. In addition, the study will provide a technical reference for the accurate and rapid measurement of seed cotton MR during harvesting operations.

Automated summary

A new method for resistive sensing of seed cotton moisture regain (MR) measurement based on pressure compensation is proposed in this study. The method was evaluated and compared with other prediction models, and the results show that it has high accuracy in predicting seed cotton MR. The study provides technical reference for the accurate and rapid measurement of seed cotton MR during harvesting operations.