Accurate yield estimation before the wheat harvest is very important for precision management, maintaining grain market stability, and ensuring national food security. In this study, to further improve the accuracy of winter wheat yield estimation, machine learning models, including GPR, SVR, and DT, were employed to construct yield estimation models based on the single and multiple growth periods, incorporating the color and multispectral vegetation indexes. The results showed the following: (1) Overall, the performance and accuracy of the yield estimation models based on machine learning were ranked as follows: GPR, SVR, DT. (2) The combination of color indexes and multispectral vegetation indexes effectively improved the yield estimation accuracy of winter wheat compared with the multispectral vegetation indexes and color indexes alone. The accuracy of the yield estimation models based on the multiple growth periods was also higher than that of the single growth period models. The model with multiple growth periods and multiple characteristics had the highest accuracy, with an R² of 0.83, an RMSE of 297.70 kg/hm², and an rRMSE of 4.69%. (3) For the single growth period, the accuracy of the yield estimation models based on the color indexes was lower than that of the yield estimation models based on the multispectral vegetation indexes. For the multiple growth periods, the accuracy of the models constructed by the two types of indexes was very close, with R² of 0.80 and 0.80, RMSE of 330.37 kg/hm² and 328.95 kg/hm², and rRMSE of 5.21% and 5.19%, respectively. This indicates that the low-cost RGB camera has good potential for crop yield estimation. Multi-temporal and multi-sensor remote sensing data fusion can further improve the accuracy of winter wheat yield estimation and provide methods and references for winter wheat yield estimation.

To establish a collision model of wheat grains impacting a force plate with a piezoelectric sensor, and to investigate the influence of the elastic recovery coefficient on the sensor’s detection accuracy during the collision process, this study employed object kinematic principles to construct a wheat elastic recovery coefficient measurement device. This device ascertains the elastic properties of wheat during collisions and determines the elastic recovery coefficient of the wheat collision model. The wheat variety Jinan No. 17 was selected for testing, and the effects of the contact material, grain drop height, material thickness, and grain moisture content on the collision recovery coefficient during the collision process were analyzed through single-factor and multi-factor experiments. The experimental results demonstrate that the collision recovery coefficient of wheat grains increases with the stiffness of the collision materials for different materials. The grain recovery coefficient of wheat exhibits a downward trend with increasing falling height and moisture content, while it tends to rise as the material thickness increases. Data analysis and comparison reveal that, given the determination of the collision material, the moisture content of wheat exerts the most significant effect on the elastic recovery coefficient, followed by material thickness, while the influence of falling height is less pronounced. The findings of this study can provide data support for simulation testing and product design of wheat combine harvester cleaning screen body mechanisms and wheat seeders.

In this study, in order to achieve accurate detection of grain loss during the process of combine harvester harvesting, a piezoelectric-based grain loss sensor has been developed. The sensor utilizes a double-layer circular piezoelectric ceramic sheet as its sensitive element. When different grains come into contact with the sensitive element, the piezoelectric effect of the ceramic sheet generates corresponding charges. These charges are then converted into knock charge signals through a charge amplification signal processing circuit that has been specially designed for this purpose. The harsh operating conditions of the sensor, including the presence of significant vibration and noise interference, necessitate the incorporation of a double-layer vibration-damping structure in both the top and bottom layers of the sensor. This paper seeks to analyze the vibration-damping effect of various shock-absorbing materials and structures incorporated into the sensor. This is accomplished by creating a dynamic analysis model that accounts for vibration interference. Furthermore, an experimental bench is established for the purpose of verifying the vibration-damping test results. These tests demonstrate that the utilization of a properly selected vibration-damping structure and materials can effectively eliminate mechanical vibration and noise interference. This, in turn, leads to improved detection accuracy of charge signals after knocking and enhances the overall anti-interference ability of the sensor.