Wheat harvesting is one of the most important links in the whole wheat production process. In China, the wheat planting areas are wide, and the patterns are diversified. In addition, the problem of harvest losses caused by the numerous brands and low performance of domestic combine harvesters has always existed. Any losses during harvesting will result in less income for the farmers. Therefore, according to the actual situation of mechanized wheat harvesting and the losses occurring within different parts of the harvester, it is of great significance to select the appropriate loss reduction methods to effectively reduce wheat harvest losses. In accordance with the problems of loss during mechanized harvesting, this research first points out the main losses in the operation of a wheat combine harvester, then introduces sensor monitoring technology for grain harvesting loss and intelligent control technology for the combine harvester and analyzes their application to loss reduction in mechanized wheat harvesting. Finally, we put forward conclusions and suggestions on this loss reduction technology for wheat mechanization harvesting in order to provide a reference for reducing the losses and promoting the sustainable development of modern agriculture.

Traditional harvesters of tiger nuts (Cyperus esculentus) face problems including low harvesting efficiency, high loss rate, high impurity rate, and high labor intensity. To solve these problems and improve the harvesting efficiency and quality of tiger nuts, a crawler-type tiger-nut combine harvester that integrates digging, soil removal, picking, screening, and collection was designed. The machinery comprises crawler devices and working devices. The key devices were designed through theoretical analysis. Therein, the digging and hoisting devices consist of digger blades, combined soil-breaking blades, and vibrating hoisting chains. The tuber picking and screening device is composed of the tuber picking drum, double-deck heterodromous vibrating screens, impurity removal blowers, and soil-crushing guide rollers. The crawler devices include the track assemblies and the hydraulic driving systems. SolidWorks was used to establish the virtual prototype model. Combined with simulation using the discrete element software, the law of motion of tiger-nut tubers in the digging, elevation, and screening processes was studied, which verified the feasibility of the design. Finally, a prototype was manufactured and fabricated to conduct field harvesting tests on tiger nuts. Field test results indicate that the harvesting efficiency, harvest rate, and impurity rate of the tiger-nut harvester are separately 0.216 ha/h, 98.14%, and 3.24%, which meet the harvesting requirements for tiger-nut growers.

Aiming at the problem of a poor separation of tiger nut, soil and grass during harvest, a hoisting device consisting of a combined-type hoisting sieve, vibrating wheels and soil roller was designed in combination with the requirements of the planting and harvesting of tiger nut. Through a theoretical analysis of the movement of the mixture of tiger nut, sand and grass in the process of hoisting, the basic law that affects the soil filter rate was determined, and the parameters affecting the soil-sieving rate were determined, and the hoisting angle, linear hoist speed and range of related parameters of vibrating wheels were obtained. Based on the DEM-MBD method, a simulation model of an excavating and hoisting device was built. With the hoisting angle, linear hoist speed, vibrating frequency and vibrating amplitudes of vibrating wheels as test factors, and the soil-sieving rate as test index, an orthogonal rotating-center combination test with four factors and three levels was carried out. The results showed that the influence of various factors on soil-sieving rate was as follows: vibrating frequency of vibrating wheels > linear hoist velocity > vibrating amplitudes of vibrating wheels > hoisting angle. When the vibrating frequency of the vibrating wheels was 9 Hz, the linear hoist speed was 0.66 m/s, vibrating amplitude of vibrating wheels was 25 mm and hoisting angle was 26°; the maximum value of the soil-sieving rate was 42.5%. The optimized parameters were applied to field test for verification, and the soil-sieving rate of the field test was 44.7%, which was better than the simulation test. The research results can provide a theoretical reference for design optimization and simulation analysis of tiger nut harvesters.

Predicting the harvest time of wheat in large areas is important for guiding the scheduling of wheat combine harvesters and reducing losses during harvest. In this study, Zhumadian, Zhengzhou and Anyang, the main winter-wheat-producing areas in Henan province, were selected as the observation points, and the main producing areas were from south to north. Based on Landsat 8 satellite remote sensing images, the changes in NDVI (Normalized Difference Vegetation Index), EVI (Enhanced Vegetation Index), and NDWI (Normalized Difference Water Index) were analyzed at different growth stages of winter wheat in 2020. Multiple regression analysis and Back Propagation (BP) neural network machine learning methods were used to establish prediction models for the harvest time of winter wheat at different growth stages. The results showed that the prediction model based on a BP neural network had high accuracy. The RMSE, MAE and MAPE of the training set and the test set were 0.531 and 0.5947, 0.3001 and 0.3104, 0.0114% and 0.0119%, respectively. The prediction model of winter wheat harvest date based on BP neural network was verified in the main winter wheat producing areas of Henan province in 2020 and 2021. The average errors were 1.67 days and 2.13 days, which were less than 3 days, meeting the needs for winter wheat production and harvest. The grain water content of winter wheat at harvest time calculated by the prediction model reached the grain water standard of the wheat combine harvester. Therefore, the prediction of the winter wheat harvest time can be realized based on multiple remote sensing indicators.

Real-time knowledge of kernel breakage during corn harvesting plays a significant role in the adjustment of operational parameters of corn kernel harvesters. (1) Transfer learning by initializing the DenseNet121 network with pre-trained weights for training and validating a dataset of corn kernels was adopted. Additionally, the feature extraction capability of DenseNet121 was improved by incorporating the attention mechanism of a Convolutional Block Attention Module (CBAM) and a Feature Pyramid Network (FPN) structure. (2) The quality of intact and broken corn kernels and their pixels were found to be coupled, and a linear regression model was established using the least squares method. The results of the test showed that: (1) The MAP^b50 and MAP^m50 of the improved Mask Region-based Convolutional Neural Network (RCNN) model were 97.62% and 98.70%, in comparison to the original Mask Region-based Convolutional Neural Network (RCNN) model, which were improved by 0.34% and 0.37%, respectively; the backbone FLOPs and Params were 3.09 GMac and 9.31 M, and the feature extraction time was 206 ms; compared to the original backbone, these were reduced by 3.87 GMac and 17.32 M, respectively. The training of the obtained prediction weights for the detection of a picture of the corn kernel took 76 ms, so compared to the Mask RCNN model, it was reduced by 375 ms; based on the concept of transfer learning, the improved Mask RCNN model converged twice as quickly with the loss function using pre-training weights than the loss function without pre-training weights during training. (2) The coefficients of determination R² of the two models, when the regression models of the pixels and the quality of intact and broken corn kernels were analyzed, were 0.958 and 0.992, respectively. These findings indicate a strong correlation between the pixel characteristics and the quality of corn kernels. The improved Mask RCNN model was used to segment mask pixels to calculate the corn kernel breakage rate. The verified error between the machine vision and the real breakage rate ranged from −0.72% to 0.65%, and the detection time of the corn kernel breakage rate was only 76 ms, which could meet the requirements for real-time detection. According to the test results, the improved Mask RCNN method had the advantages of a fast detection speed and high accuracy, and can be used as a data basis for adjusting the operation parameters of corn kernel harvesters.