In view of the problems of a poor seed-filling state and poor seed-cleaning effect of a precision metering device under high-speed working conditions, an inside-filling air-assisted high-speed precision maize seed-metering device was designed, fabricated, and evaluated. The overall structure and working principles of the seed-metering device are explained. Theoretical analysis and parametric design of the key parts, such as the horizontal stirring notch seed-metering plate and the wheel seed-cleaning mechanism, were carried out. The main factors that produce multiple seeding problems were studied. The forward speed (FS) and working pressure (WP) of the seed-metering device are determined as the main factors affecting the seeding performance. A quadratic orthogonal rotation combination experiment was designed. The relationship between influencing factors and performance indexes was constructed using the Design-Expert 10.0.7 software. The response surface method was used to optimize the experimental results. The results of the bench validation experiments showed that the seeding performance of the seed-metering device was excellent when the FS of the seed-metering device was 13.10 km/h and the WP of the seed-metering device was 4.75 kPa. The results of the bench comparison showed that the seeding performance of the designed seed-metering device was better than that of the conventional air-suction precision seed-metering device. This study provides theoretical and practical references for the optimized design of a pneumatic precision seed-metering device and the sustainable development of grain.

The prediction of crop yield plays a crucial role in national economic development, encompassing grain storage, processing, and grain price trends. Employing multiple sensors to acquire remote sensing data and utilizing machine learning algorithms can enable accurate, fast, and nondestructive yield prediction for maize crops. However, current research heavily relies on single-type remote sensing data and traditional machine learning methods, resulting in the limited robustness of yield prediction models. To address these limitations, this study introduces a field-scale maize yield prediction model named the convolutional neural network–attention–long short-term memory network (CNN-attention-LSTM) model, which utilizes multimodal remote sensing data collected by multispectral and light detection and ranging (LIDAR) sensors mounted on unmanned aerial vehicles (UAVs). The model incorporates meteorological data throughout the crop reproductive stages and employs the normalized difference vegetation index (NDVI), normalized difference red edge (NDRE), soil-adjusted vegetation index (SAVI), and enhanced vegetation index (EVI) for the initial part of the vegetative stage (initial part of the V period), the later part of the vegetative stage (later part of the V period), the reproductive stage (R period), and the maturity stage (M period), along with LIDAR data for Point_75–100 in the later part of the V period, Point_80–100 in the R period, and Point_50–100 in the M period, complemented by corresponding meteorological data as inputs. The resulting yield estimation demonstrates exceptional performance, with an R² value of 0.78 and an rRMSE of 8.27%. These results surpass previous research and validate the effectiveness of multimodal data in enhancing yield prediction models. Furthermore, to assess the superiority of the proposed model, four machine learning algorithms—multiple linear regression (MLR), random forest regression (RF), support vector machine (SVM), and backpropagation (BP)—are compared to the CNN-attention-LSTM model through experimental analysis. The outcomes indicate that all alternative models exhibit inferior prediction accuracy compared to the CNN-attention-LSTM model. Across the test dataset within the study area, the R² values for various nitrogen fertilizer levels consistently exceed 0.75, illustrating the robustness of the proposed model. This study introduces a novel approach for assessing maize crop yield and provides valuable insights for estimating the yield of other crops.

In order to accurately establish a discrete element model for the whole plant flexibility of upright rice during the harvesting period, several physical characteristics, such as geometric features, moisture content, and density, of the entire rice plant were measured, along with frictional properties, such as the static and rolling friction coefficients, and mechanical properties, including the elastic modulus and restitution coefficient. A flexible and upright discrete element model of the rice plant was established using the DEM method based on the Hertz–Mindlin (no slip) and Hertz–Mindlin with bonding mechanical models. The parameters were optimized through Plackett–Burman screening experiments, steepest ascent experiments, and Box–Behnken optimization experiments to accurately determine the discrete element model parameters of each component of the rice plant. The calibration process of the contact parameters between rice grains and steel was analyzed in detail as an example, resulting in a calibration error of 0.68% for the natural repose angle. Taking the calibration of the contact parameters between the main stem and steel as an example, a detailed analysis of the calibration process was conducted. The calibration resulted in a calibration error of 2.76% for the natural repose angle and 2.33% for deflection. This study lays the foundation for understanding the mechanical response of rice and machinery when they are coupled together. Additionally, it provides valuable references for establishing discrete element models of plant species other than rice.

The physical properties of maize seeds are closely related to food processing and production. To study and evaluate the characteristics of maize seeds, typical maize seeds in a cold region of North China were used as test varieties. A variety of agricultural material test benches were built to measure the maize seeds’ physical parameters, such as thousand-grain weight, moisture content, triaxial arithmetic mean particle size, coefficient of static friction, coefficient of rolling friction, angle of natural repose, coefficient of restitution, and stiffness coefficient. Principal component and cluster comprehensive analyses were used to simplify the characteristic parameter index used to judge the comprehensive score of maize seeds. The results showed that there were significant differences in the main physical characteristics parameters of the typical maize varieties in this cold area, and there were different degrees of correlation among the physical characteristics. Principal component analysis was used to extract the first three principal component factors, whose cumulative contribution rate was over 80%, representing most of the information of the original eight physical characteristic parameters, and had good representativeness and objectivity. According to the test results, the classification standard of the evaluation of the physical characteristics of 15 kinds of maize seeds were determined, and appropriate evaluations were conducted. The 15 kinds of maize seeds were clustered into four groups by cluster analysis, and the physical characteristics of each groups were different. This study provides a new idea for the evaluation and analysis of the physical properties of agricultural materials, and provides a new method for the screening and classification of food processing raw materials.

This paper aims to solve the problem of high reseeding rates and mis-seeding rates in the rice multi-grain hole direct seeding process. A multi grain cluster air suction type rice hill direct seed metering device was developed, and its seeding mechanism was analyzed. Based on CFD-DEM coupling simulation and bench tests, this study explored and optimized the performance of the seed metering device, and carried out the seeding adaptability test. The simulation results were as follows: when the negative pressure was −5 kPa, the static pressure, dynamic pressure and velocity of the flow field reached the maximum. When the negative pressure was −4 kPa, the qualification index was 89.62%, the reseeding index was 4.36%, and the mis-seeding index was 6.02%. The results of the orthogonal rotation combination test of three factors and five levels showed that when the rotation speed, negative pressure and the length of stirring brush were 20.70 rpm, −4.0 kPa and 10.50 mm respectively, the seed metering performance was the best, the qualification index was 90.85%, the reseeding index was 4.41% and the mis-seeding index was 4.74%. The seed metering device had strong adaptability to the sowing of different rice varieties, and met the agronomic requirements of direct seeding and seeding in rice holes.