Green forage maize harvesters face challenges such as high soil humidity and soft soil in the field, mismatched working parameters, and poor reliability and adaptability. These challenges often result in header blockage, significant harvest loss, and increased energy consumption. Traditional testing and statistical analysis methods used in most existing studies are limited by complex test processes, their time-consuming nature, high costs, and poor prediction accuracy. To address these problems, a test bench was constructed to analyze the effects of forward speed, cutting height, number of rows, and their interactions on specific energy consumption and harvest loss of the green forage maize (GFM) header. A combined response surface method (RSM)–artificial neural network (ANN) approach is proposed for modeling and predicting the performance parameters of the header. The optimal conditions were determined by optimizing the specific energy consumption and loss rate. The optimal combination parameters are a forward speed of 1.6 km/h, a cutting height of 167 mm, and a number of rows of 4. However, RSM–ANN has larger R² values and lower root mean square errors (RMSE) and mean square errors (MSE) compared to RSM. Specifically, the R² of the RSM–ANN model for specific energy consumption and loss rate a 0.9925 and 0.9906, MSE are 0.00001775 and 0.004558, and RMSE are 0.004214 and 0.006752, respectively. The results show that the combined RSM–ANN method has higher precision and accuracy and can better predict and optimize the header performance. This study overcomes the limitations of traditional methods and has the potential to provide data and method references for the design, optimization, prediction, and intelligent diagnosis of faults in the operational parameters of agricultural machinery.

Stalk lodging is prone to occur during the harvesting of green forage corn due to head and root anchorage, resulting in loss of harvest. In particular, the fallen stalk wraps around the head, causing a blockage and increasing the energy cost. To address this problem, the deflection model was analyzed and established and a novel method to explore the field bend characteristics of green forage maize was proposed. The effects of stalk diameter, bending angle, and cutting height on bending stress and Young’s modulus were explored by using the method of in-situ measurement. Additionally, the bending deflection, axial displacement and measuring point displacement were obtained. Experimental results indicate that the stalk diameter and bending angle have a significant influence on bending stress. The bending stress has a positive correlation with bending angle and cutting height, while there is a negative correlation with stalk diameter. On the other hand, the bending angle, stalk diameter, and cutting height are closely related to the Young’s modulus. The mean values of the Young’s modulus decrease as a quadratic function with the increasing diameter and bending angle, while the cutting height has the opposite effect on it. Besides, the average values of bending deflection, axial displacement, and measuring point displacement exhibit an increasing trend when the bending angle and cutting height increase. This study results can provide a reference for studying the failure mechanism of green forage maize stalk harvesting, and the design of green forage maize harvesting machinery. Keywords: green forage maize, field measurement, bending stress, Young’s modulus, deflection model, harvest DOI: 10.25165/j.ijabe.20231604.8043 Citation: Fu J, Xue Z, Fu Q K, Chen Z, Ren L Q. Bending properties of green forage maize in field. Int J Agric & Biol Eng, 2023; 16(4): 51–57.

Image-based measurement of chopped maize stalk length remains difficult because the fragments are often slender, curved, touching, or partly overlapped. Bounding-box dimensions are therefore not reliable for length estimation, and manual measurement is too slow for repeated quality assessment. In this study, we developed a YOLO11n-based visual framework for measuring chopped maize stalk length under fixed imaging conditions. The dataset contained 1127 images collected on a laboratory platform and covered stalk lengths of 10–150 mm, different moisture states, and isolated, touching, and overlapping arrangements. To obtain more stable regions of interest, the YOLO11n detector was modified with large separable kernel attention (LSKA), a lightweight cross-scale decoupled detection (LSCD) head, and Wise intersection over union version 3 (WIoU v3). The detected stalk regions were then processed by local segmentation, morphological refinement, skeleton extraction, longest-path calculation, and washer-based scale conversion. The modified detector reached 94.8% precision, 90.4% recall, 96.5% mAP@0.5, and 71.1% mAP@0.5:0.95, with a detector inference speed of 174 FPS. In the length-measurement test, the mean relative errors were 5.8%, 8.3%, and 10.4% for the 80 mm groups, respectively. Across all evaluated fragments, the complete pipeline produced an MAE of 6.0 mm, an RMSE of 9.4 mm, and a mean relative error of 8.2%. The framework therefore provides a practical way to measure chopped maize stalk length under controlled imaging conditions, although long, curved, and cluttered fragments still caused most of the remaining errors.