This paper provides a review of path-tracking strategies used in autonomous agricultural vehicles, mainly from two aspects: vehicle model construction and the development and improvement of path-tracking algorithms. Vehicle models are grouped into numerous types based on the structural characteristics and working conditions, including wheeled tractors, tracked tractors, rice transplanters, high clearance sprays, agricultural robots, agricultural tractor–trailers, etc. The application and improvement of path-tracking control methods are summarized based on the different working scenes and types of agricultural machinery. This study explores each of these methods in terms of accuracy, stability, robustness, and disadvantages/advantages. The main challenges in the field of agricultural vehicle path tracking control are defined, and future research directions are offered based on critical reviews. This review aims to provide a reference for determining which controllers to use in path-tracking control development for an autonomous agricultural vehicle.

Straw mulching in conservation tillage farmland can effectively promote land utilization and conservation. However, in this farming mode, surface straw suppresses weed growth, affecting weed size and position distribution and obscuring the weeds, which hampers effective weed management in the field. Accurate weed identification and localization, along with efficient herbicide application, are crucial for achieving precise, efficient, and intelligent precision agriculture. To address these challenges, this study proposes a weed detection model for a targeted spraying system. Firstly, we collected the dataset of weeds in a straw-covered environment. Secondly, we proposed an improved YOLO v5s network, incorporating a Convolutional Block Attention Module (CBAM), FasterNet feature extraction network, and a loss function to optimize the network structure and training strategy. Thirdly, we designed a targeted spraying system by combining the proposed model with the targeted spraying device. Through model test and spraying experiments, the results demonstrated that while the model exhibited a 0.9% decrease in average detection accuracy for weeds, it achieved an 8.46% increase in detection speed, with model memory and computational load reduced by 50.36% and 53.16%, respectively. In the spraying experiments, the proposed method achieved a weed identification accuracy of 90%, a target localization error within 4%, an effective spraying rate of 96.3%, a missed spraying rate of 13.3%, and an erroneous spraying rate of 3.7%. These results confirm the robustness of the model and the feasibility of the targeted spraying method. This approach also promotes the application of deep learning algorithms in precision weed management within directional spraying systems.

The shape of particles is a critical determinant that significantly influences the accuracy of discrete element simulations. To reduce the discrepancies between the discrete element model of wheat seeds and the actual particle shapes, and to enhance the accuracy of Computational Fluid Dynamics-Discrete Element Method (CFD-DEM) coupling simulations in gas–solid two-phase flow studies, We employed laser scanning and inverse modeling techniques to develop a three-dimensional (3D) reconstruction of the wheat seed. Subsequently, we employed Rocky DEM simulation software to develop a polyhedron model and an Angle of Repose (AOR) test model. The interval range of material parameters was determined through a series of physical experiments and subsequently employed to delineate the high and low levels of parameters for the simulation tests. The simulation parameters were calibrated using data from AOR simulation tests. The Plackett–Burman test, Steepest-Ascent test, and Box–Behnken test were conducted sequentially to determine the optimal parameter configuration. A test bench for wheat gas-assisted seeding was constructed, and a semi-resolved CFD-DEM coupling simulation model was developed to perform comparative analysis. The results demonstrated that the optimal parameters were as follows: the static friction coefficient of wheat seed was 0.15, the dynamic friction coefficient of wheat seed was 0.11694, and the dynamic friction coefficient between wheat seed and resin was 0.0797. In this scenario, the relative error of AOR was 2.3% and the maximum relative error of ejection velocity observed was 4.1%. The reliability of the polyhedron model and its calibration parameters was rigorously validated, thereby providing a robust reference for studies on gas–solid two-phase flows.

In arid regions characterized by extreme water scarcity, it is important to synergistically optimize both crop yield and water use. Irrigation strategies based on empirical knowledge overlook crops’ dynamic water needs and may cause water waste and yield loss. To address this issue, this paper proposes an intelligent irrigation scheduling method based on a crop growth model and an improved deep reinforcement learning (DRL) agent. We construct a high-fidelity cotton growth environment using the Decision Support System for Agrotechnology Transfer (DSSAT) model. The model was calibrated with local data from the Shihezi region, Xinjiang, to provide a reliable simulation platform for DRL agent training. We developed a temporal state representation module based on a Bidirectional Long Short-Term Memory (BiLSTM) network and an attention mechanism. This module captures dynamic trends in historical environmental information to focus on critical decision factors. The Soft Actor–Critic (SAC) algorithm was improved by integrating a feature attention mechanism to enhance decision-making precision. A dynamic reward function was designed based on the critical growth stages of cotton to incorporate agronomic prior knowledge into the optimization objective. Simulation results demonstrate that our proposed method can improve water use efficiency (WUE) by 39.0% (with an 8.4% increase in yield and a 22.1% reduction in water consumption) compared to fixed-schedule irrigation strategies. An ablation study further confirms that each of our proposed modules—BiLSTM, the attention mechanism, and the dynamic reward—makes a significant contribution to the final performance.

In rice–wheat rotation areas of China, traditional wheat seeders have severe blockage, low working efficiency and poor seeding quality. In this study, a pneumatic shooting technology was designed, consisting mainly of a nozzle, shell and acceleration tube. To improve the sowing depth of the pneumatic shooting device, the response-surface methodology of structure parameters and CFD simulation technology was adopted in this work. The effects of working pressure, acceleration-tube diameter and throat distance on the steady airflow length (SAL) and steady airflow velocity (SAV) were studied by airflow field analysis, and the response-surface method was introduced to obtain the optimal parameter combination of the pneumatic shooting device for wheat. The optimal parameter combination was working pressure 686 kPa, acceleration tube diameter 8 mm and throat distance 20 mm. The simulation result showed that the optimized device of pneumatic shooting performs faster and has more stable airflow field characteristics in comparison to the initial device. The field test demonstrated that the optimized device has about 68% higher seeding depth than the initial device. The average field-seeding depth of the optimized device was 19.95 mm, which is about 68% higher than the initial device. The emergence rate for the optimized device was about 88.7% without obvious reduction. CFD and response-surface methods positively affect the optimization of pneumatic wheat-shooting devices, and the significance was also confirmed.

The article briefly describes the importance of furrowing depth stability for seed germination and growth under precision seeding conditions. By analyzing the application status of furrowing depth control technology of the planter globally, the research method, technical characteristics, and development of furrowing depth stability control technology are reviewed from three key aspects, namely, profiling adjustment device, furrowing depth detection technology, and automatic control system. In this paper, (1) two types of profiling adjustments, active and passive, are described based on the difference in the downforce adjustment method; (2) three furrowing depth detection methods are described based on different sensors; (3) and three ways of regulating the furrowing depth system are summarized based on the different ways of evaluating the stability of furrowing depth. In addition, the characteristics and application requirements of global furrow depth control technology are summarized. It is proposed that the future planter should be developed in the direction of automatic navigation, automatic monitoring and evaluation of seeding quality, variable seeding, high-speed seeding, and other intelligent precision seeding techniques. The summary and outlook of this paper aim to promote the overall development of furrowing depth control technology.