The air-suction precision seed-metering device is prone to the instability of the seed adsorption state, which arises from blockage of the suction hole and leads to uneven seeding. This paper analyzed and determined key structural parameters of the seed-metering plate, then established an adsorption mechanics model of the seed during the migration process and designed the key structure of the air-suction seed-metering device with the aim of improving the uniformity of high-speed direct seeding of vegetables. Furthermore, we used the DEM-CFD coupling method to analyze the influence of the law of seeds on the change of the flow field with different hole types. Results showed that the turbulent kinetic energy (202.65 m²∙s⁻²) and the coupling force to the seeds (0.029 N) of the B-type hole are the largest, which is the best fluid domain structure for the suction hole of the seed-metering plate. Moreover, we used Adams to analyze the meshing process between the knock-out wheel and the seed-metering plate, affirming the rationality of the knock-out wheel design. Finally, in order to improve the working efficiency of the seed-metering device, we performed one-factor and response surface experiments of seeding performance using the air-suction seed-metering device designed with the optimized structure as the experimental object. Analysis of the influence of weights across each factor on the experimental performance evaluation indicators revealed an optimal combination of seeding performance parameters in the air-suction seed-metering device, namely a seed-throwing angle of 13°, a working speed of 14.5 km/h, and negative pressure of 3.1 kPa. Results from verification experiments revealed the corresponding experimental indicators, namely qualified, multiple, and missing indexes of 95.9, 1.2%, and 2.9%, respectively.

Laser-induced graphene (LIG) has been widely used in flexible sensors due to its excellent mechanical properties and high conductivity. In this paper, a flexible pressure sensor prepared by bionic micro/nanostructure design and LIG mass fraction regulation is reported. First, prepared LIG and conductive carbon paste (CCP) solutions were mixed to obtain a conductive polymer. After the taro leaf structure was etched on the surface of the aluminum alloy plate by Nd:YAG laser processing, the conductive polymer was evenly coated on the template. Pressure sensors were packaged with a stencil transfer printing combined with an Ecoflex flexible substrate. Finally, the effects of different laser flux and the proportion of LIG in the composite on the sensitivity of the sensor are discussed. The results show that when the laser flux is 71.66 J·cm⁻² and the mass fraction of LIG is 5%, the sensor has the best response characteristics, with a response time and a recovery time of 86 ms and 101 ms, respectively, with a sensitivity of 1.2 kPa⁻¹ over a pressure range of 0–6 kPa, and stability of 650 cycle tests. The LIG/CCP sensor with a bionic structure demonstrates its potential in wearable devices.

Tea is a popular beverage worldwide and also has great medical value. A fundamental understanding of tea shoot growth and a precision picking model should be established to realize mechanized picking of tea shoots with a small product loss. Accordingly, the terminal bud length (L_bud), tea stem length (L_stem), terminal bud angle (α_bud), tea stem angle (α_stem), and growth time (t) were considered as the key growth parameters; the sum of the vertical lengths of the terminal bud and stem (ξ), the picking radius (r), and the vertical length of the stem (Z_stem) were considered as the picking indexes of the tea shoots. The variations in growth parameters with time were investigated using a 3-D coordinate instrument, and the relationships between the growth parameters and the picking indexes were established using an artificial neural network (ANN). The results indicated that the tea growth cycles for periods P₁, P₂, P₃, P₄, P₅, and P₆ were 14, 7, 6, 4, 4, and 6 d, respectively. A growth cycle diagram of the tea growth was established. Moreover, a 5-2-12-3 ANN model was developed. The best prediction of ξ, r, and Z_stem was found with 16 training epochs. The MSE value was 0.0923 × 10⁻⁴, and the R values for the training, test, and validation data were 0.99976, 0.99871, and 0.99857, respectively, indicating that the established ANN model demonstrates excellent performance in predicting the picking indexes of tea shoots.

In order to precisely reproduce the precise seeding process of the population in the air-suction seed-metering device, it is necessary to execute accurate modeling of seed particles using the bonded-particle model, in combination with the discrete element method (DEM) and computational fluid dynamics (CFD). Through the repose angle, slope screening, rotating container, and particle sedimentation experiments, in this paper, the influence of the filling accuracy of the bonded-particle model on the flow behavior and mixing characteristics of the seed population was first explored based on EDEM software. The viability of the suggested modeling approach for pelleted vegetable seeds, as described in this study, was confirmed by comparing experimental and simulation outcomes. The surface roughness values obtained from the studies above were utilized to assess the accuracy of the bonded-particle model in filling. Additionally, a mathematical technique for determining the surface roughness was provided. Furthermore, an analysis of the multiple contacts in the bonded-particle model was also performed. The results indicated that the simulation results closely matched the experimental data when the number of sub-spheres in the bonded-particle model was equal to or more than 70, as measured by the standard deviation. In addition, the most optimal modeling scheme for the pelletized vegetable seed bonded-particles, based on the cost of coupling simulation, was found to be the bonded-particle surface roughness (BS) with a value of 0.1. Ultimately, a practical example was utilized to demonstrate the utilization of the pelleted vegetable seed bonded-particle model and the DEM-CFD coupling approach in analyzing the accuracy of the seeding process in the air-suction seed-metering device. This example will serve as a valuable reference point for future field studies.

In unstructured tea garden environments, accurate recognition and pose estimation of tea bud leaves are critical for autonomous harvesting robots. Due to variations in imaging distance, tea bud leaves exhibit diverse scale and pose characteristics in camera views, which significantly complicates the recognition and pose estimation process. This study proposes a method using an RGB-D camera for precise recognition and pose estimation of tea bud leaves. The approach first constructs an for tea bud leaves, followed by a dynamic weight estimation strategy to achieve adaptive pose estimation. Quantitative experiments demonstrate that the instance segmentation model achieves an mAP@50 of 92.0% for box detection and 91.9% for mask detection, improving by 3.2% and 3.4%, respectively, compared to the YOLOv8s-seg instance segmentation model. The pose estimation results indicate a maximum angular error of 7.76°, a mean angular error of 3.41°, a median angular error of 3.69°, and a median absolute deviation of 1.42°. The corresponding distance errors are 8.60 mm, 2.83 mm, 2.57 mm, and 0.81 mm, further confirming the accuracy and robustness of the proposed method. These results indicate that the proposed method can be applied in unstructured tea garden environments for non-destructive and precise harvesting with autonomous tea bud-leave harvesting robots.

This article proposes a method for compiling the load spectra of reducers for hybrid electric vehicles. Selecting typical working conditions for real vehicle data collection, the load data under each typical working condition were divided into five categories according to the state of the power source and the data were preprocessed. The optimal sample loads for compiling load spectra were obtained based on a multi-criteria decision-making method, rainflow counting for optimal sample loads was performed according to different power source output patterns, non-parametric extrapolation was performed to obtain the full-life two-dimensional load spectrum after dimensionality reduction, and a full-life eight-level programmed load spectrum that could be used for bench tests was obtained. Using the programmed load spectrum and the extracted sample load as the load input, a fatigue life prediction simulation of the reducer gear of a hybrid electric vehicle was carried out. The reducer gear fatigue life from the programmed load spectrum was compared to the gear fatigue life under actual load. The fatigue life of the reducer gear when the programmed load spectrum was used as the input was 1.412 × 10³. When the actual load was used as the input load, the fatigue life of the reducer gear was 1.933 × 10³. The relative error between the two is only 26%, which is in the normal range. The results show that the programmed load spectrum is effective and reliable and that the load spectrum compilation method provides a basis for accurately evaluating the reliability of the hybrid electric vehicle reducer.