The cleaning device is an important part of combine harvesters, as its superior or inferior performance directly affects the performance of the combine harvester greatly. With an increasing rice yield, the current single-duct cleaning performance declines greatly, and causes a large grain sieve loss level and a direct grain loss for the farms. To optimize the existing single-duct cleaning device to meet the large feeding rate requirement, firstly, the terminal velocity of rice grain and MOG (material other than grain) for different varieties was experimentally measured by the custom-made device. The effects of the moisture content of rice grains and the length of short straws on terminal velocity were studied in detail. Then, the gas–solid two-phase flow theory was comprehensively applied by utilizing the discrete element method (DEM) and computational fluid dynamics (CFD) to study the working mechanism of the existing single-duct cleaning unit, and the cleaning performance was evaluated from the view of the motion law of the threshing output within the cleaning shoe. At last, a multi-duct cleaning device was put forward, and a field experiment was performed to assess the performance of the newly developed cleaning device. The results showed that the grain sieve losses ratio and grain impurity ratio improved dramatically, proving that structure optimization of the cleaning device was feasible.

Feeding rates serve as a vital indicator for adjusting the working parameters of the combine harvester. A non-invasive diagnostic approach to predicting the feed rates of combine harvesters by collecting vibration signals of the inclined conveyor was introduced in this study. To establish a feed rate prediction model, the correlation between feeding rates and vibration signal characteristics was investigated. Vibration signal characteristics in both the time domain and frequency domain were also analyzed in detail. The RMS (root mean square) value and the total RMS value of the one-third octave extracted from the vibration signal were utilized to establish a feed rate prediction model, and field tests were conducted to verify the model performance. The experimental results indicated that the relative errors of the established model range from 3.1% to 4.9% when harvesting rice. With the developed feed rate prediction system, the control system of the combine harvester can acquire feed rate information in real time, and the working parameters can be adjusted in advance, thereby, it can be expected to greatly enhance the working performance of the combine harvesters.