Search

To improve the automation level and operation quality of China's beet harvester and reduce the loss due to damaged and missed excavation, this study used a self-developed sugar beet combine harvester and field simulation experiment platform, based on the single-factor bench test of the automatic row following system in the early stage, taking hydraulic flow A, spring preload B, and forward speed C which have significant influence on performance indices as test factors, and taking the missed excavation rate, breakage rate and reaction time as performance indices, the orthogonal experimental study on the parameter optimization of the three-factor and three-level automatic row following system with the first-order interaction of various factors was carried out. The results of the orthogonal experiments were analyzed using range analysis and variance analysis. The results showed that there were differences in the influence degree, factor priority order and first-order interaction, and the optimal parameter combination on each performance index. A weighted comprehensive scoring method was used to optimize and analyze each index. The optimal parameter combination of the overall operating performance of the automatic row following system was A2B2C1, that is, the hydraulic flow was 25 L/min, the forward speed was 0.8 m/s, and the spring preload was 198 N. Under this combination, the response time was 0.496 s, the missed excavation rate was 2.35%, the breakage rate was 3.65%, and the operation quality was relatively good, which can meet the harvest requirements. The comprehensive optimization results were verified by field experiments with different ridge shapes and different planting patterns. The results showed that the mean values of the missed excavation rate of different planting patterns of conventional straight ridges and extremely large "S" ridges were 2.23% and 2.69%, respectively, and the maximum values were 2.39% and 2.98%, respectively; the average damage rates were 3.38% and 4.14%, and the maximum values were 3.58% and 4.48%, which meet the industry standards of sugar beet harvester operation quality. The overall adaptability of the automatic row following system is good. This study can provide a reference for research on automatic row following harvesting systems of sugar beets and other subsoil crop harvesters. [ABSTRACT FROM AUTHOR]

A row-following system based on end-to-end learning for an agricultural robot in an apple orchard was developed in this study. Instead of dividing the navigation into multiple traditional subtasks, the designed end-to-end learning method maps images from the camera directly to driving commands, which reduces the complexity of the navigation system. A sample collection method for network training was also proposed, by which the robot could automatically drive and collect data without an operator or remote control. No hand labeling of training samples is required. To improve the network generalization, methods such as batch normalization, dropout, data augmentation, and 10-fold cross-validation were adopted. In addition, internal representations of the network were analyzed, and row-following tests were carried out. Test results showed that the visual navigation system based on end-to-end learning could guide the robot by adjusting its posture according to different scenarios and successfully passing through the tree rows. [ABSTRACT FROM AUTHOR]

A large amount of traditional fertilizer has been applied for the low utilization rate, even for the soil pollution during the current wide row-spacing of crop planting in the arid area of Northwest China. In this study, a layered deep fertilizing was proposed for the cotton planting mode (where the values of narrow and wide row spacing were 10 and 66cm, respectively), and the corn planting mode (where the values of narrow and wide row spacing were 40 and 60 cm, respectively). A series of field tests were performed on the layered deep application of base fertilizer in the direction of crop seed row using Beidou navigation precise positioning. The layered fertilization of cotton adopted the mode of a narrow row, two rows of seeds corresponding to one row of fertilizer, whereas, the corn used the mode of one row of seeds corresponding to one row of fertilizer. As such, the layered deep and row fertilization modes were fully combined during this time. The fertilizer application ratio of the upper and lower layers was 4:6, whereas the fertilization depths of the first and second layers were 10-13 cm, and 18-21 cm, respectively. The fertilizer applicator was composed of depth limiting walking wheel, frame, profiling monomer, fertilizer box, trenching layered fertilizer shovel, soil covering, and pressing device. The profiling monomer was designed to analyze the motion and mechanical characteristics. The profiling spring was then determined, according to the requirements for the quality of the profiling monomer. A bionic trenching shovel was developed to select the angular protrusion of the head of the male dung beetle. The contour curve fitting equation was obtained to optimize the structure and working parameters in the machine. The performance of soil penetration, trenching resistance, and soil disturbance were also evaluated for the bionic shovel. A soil tank test was conducted to obtain the best front edge angle of 65°, the disturbance to the soil was significantly reduced, compared with the traditional. The test results show that the trenching resistance and disturbance to the soil of the bionic shovel were significantly less than those of the traditional one, indicating a better performance of drag reduction. The split fertilizer box and layered fertilization components were designed for the machine. The upper and lower discharge ports of fertilizer were connected with the different discharge wheels in the fertilizer box. A controlling system was adopted for the working length of the groove wheel, in order to real-time adjust the fertilizer discharge wheel corresponding to the left and right fertilizer boxes. The field experiment of the prototype showed that the average deviation of row accuracy was 2.7 cm during layered fertilization, and the average depths of shallow and deep fertilizer were 12.44 and 20.49 cm for the six groups of fertilization shovel, respectively, where the consistency coefficients of variation were 6.12% and 4.98%, respectively. The comparative experiment of fertilization patterns presented the higher growth of plants and the better root development. The fertilizer application was reduced by 25% than before, while, the yields of cotton and corn increased by 8.9%, and 8.6%, respectively. Therefore, the deep fertilization of rows can be widely expected to achieve better performance in weight loss and efficiency. [ABSTRACT FROM AUTHOR]

Detection navigation baseline is primary for the automation of maize weeder in seedling. In the navigation technology based on machine vision, maize seeding or weed near the camera is photographed as a discrete area, while a plant far away from the camera is photographed as a strip area along with other plants in the same row. The two problems cannot be solved by one method. However, in this paper, an algorithm of detection navigation baseline in the row-following operation of maize weeder based on axis extraction was proposed to solve the both problems. Firstly, plants are distinguished from the background based on color feature, and the binary image is acquired. Secondly, plants are described as a set of connected components with numbers after connected components labeling and noise clearing. Thirdly, the axes of all connected components are extracted according to the calculation method of rotary inertia in physics. Next, the abnormal connected components with axes are deleted because the angles between the axes and X-axis are above angle threshold. Then, the judgment model is built based on angle tolerance and distance tolerance, the connected components in a same row based on this model through two-step traversal are merged, and a new axis is re-extracted as the axis of the plant row. Finally, the navigation baselines are detected based on the axes of the plant row. The experimental results show that the accuracy of this algorithm is more than 93%, and the computing time is less than 1.6 s, which can meet the accuracy and real-time performance requirements of maize weeder. [ABSTRACT FROM AUTHOR]