Gully erosion is one of the significant environmental issues facing the black soil regions in Northeast China, and its formation is closely related to various environmental factors. This study employs multiple machine learning models to assess gully erosion susceptibility in this region. The primary objective is to evaluate and optimize the top-performing model under high-resolution UAV data conditions, utilize the optimized best model to identify key factors influencing the occurrence of gully erosion from 11 variables, and generate a local gully erosion susceptibility map. Using 0.2 m resolution DEM and DOM data obtained from high-resolution UAVs, 2,554,138 pixels from 64 gully and 64 non-gully plots were analyzed and compiled into the research dataset. Twelve models, including Logistic Regression, K-Nearest Neighbors, Classification and Regression Trees, Random Forest, Boosted Regression Trees, Adaptive Boosting, Extreme Gradient Boosting, an Artificial Neural Network, a Convolutional Neural Network, as well as optimized XGBOOST, a CNN with a Multi-Head Attention mechanism, and an ANN with a Multi-Head Attention Mechanism, were utilized to evaluate gully erosion susceptibility in the Dahewan area. The performance of each model was evaluated using ROC curves, and the model fitting performance and robustness were validated through Accuracy and Cohen’s Kappa statistics, as well as RMSE and MAE indicators. The optimized XGBOOST model achieved the highest performance with an AUC-ROC of 0.9909, and through SHAP analysis, we identified roughness as the most significant factor affecting local gully erosion, with a relative importance of 0.277195. Additionally, the Gully Erosion Susceptibility Map generated by the optimized XGBOOST model illustrated the distribution of local gully erosion risks.

Autonomous intelligent agricultural vehicles significantly reduce labor requirements in large-scale farming operations. While real-time precise positioning of agricultural vehicles is prerequisite for autonomous operation, the global positioning system (GPS)-based positioning systems demonstrate insufficient accuracy in large-scale farming scenarios. Moreover, the limited infrastructure deployment and compromised network quality in extensive agricultural fields render conventional assistance positioning frameworks inadequate. Unmanned aerial vehicles (UAVs), as integral components of vehicular networks, facilitate positioning and navigation within space–air–ground integrated networks (SAGINs). Based on this, this article proposes a multi-UAV rapid positioning assistance architecture for large-scale farming operations. Under this architecture, local area networks (LANs) are established between UAVs and agricultural vehicles, with each vehicle equipped with a uniform linear array (ULA). The system requires only signal transmission from UAVs, while the agricultural vehicles’ local processing systems perform direction of arrival (DOA) estimation, followed by joint estimation to determine vehicle positioning coordinates. However, limited signal snapshots in practical DOA estimation environments compromise estimation accuracy. While compressed sensing (CS) algorithms demonstrate significant potential for DOA estimation applications, conventional CS methodologies employing uniform overcomplete dictionariess (UODs) in sparse representation phase incur substantial computational complexity. Additionally, the orthogonal matching pursuit (OMP) algorithm and its derivatives exhibit susceptibility to local optima convergence during signal reconstruction. Consequently, this article proposes the NODGA-CS model, integrating nonuniform overcomplete dictionary (NOD) with genetic algorithm (GA) optimization for enhanced CS-DOA estimation. Experimental results demonstrate that the proposed assistance architecture achieves superior positioning accuracy when benchmarked against state-of-the-art positioning frameworks.

The advancement of precision agriculture and smart tillage relies on high-precision, real-time perception of unmanned ground vehicle (UGV) positions. In large-scale farmland operations, conventional Global Navigation Satellite System (GNSS)-based positioning may suffer from short-term signal loss, degrading accuracy. Multi-unmanned aerial vehicle (UAV)-assisted vehicle localization based on direction-of-arrival (DOA) estimation can provide critical positioning compensation for UGV, where compressed sensing (CS)-based DOA-assisted localization algorithms are commonly employed. However, existing schemes neither account for the bias induced by the local positional oscillation of UAVs, nor address the limited accuracy and real-time performance of CS-based DOA estimation, restricting their agricultural deployment. To this end, this paper first develops an assisted-localization architecture that explicitly incorporates the local positional offsets of multiple UAVs, together with a corresponding array signal reception model. To overcome the accuracy–efficiency trade-off of conventional CS-DOA methods, an adaptive local overcomplete dictionary (LOD) is then constructed to robustly refine the angular resolution around the region of interest. With the number of sources K assumed to be known and fixed, a particle swarm optimization (PSO)-based local refinement algorithm is further introduced to adaptively optimize the DOA estimates within the constructed local dictionary, thereby improving estimation robustness under low-SNR and coherent-source conditions. Consequently, the proposed method improves robustness while maintaining favorable localization accuracy and computational efficiency in the simulated scenarios. Simulation results show that it substantially reduces localization error compared with state-of-the-art algorithms, suggesting its potential as a localization-assistance approach for UGV navigation in sustainable tillage.

The operation performance of agricultural robots is a key factor in the development of smart agriculture, while the interference of dynamic lighting environment is one of the main bottlenecks restricting its performance improvement. To address this problem, this study proposes an agricultural robot posture coupling perception operation system for dynamic lighting environment. Firstly, by analyzing the impact of different lighting conditions on the operation performance of agricultural robots, a light coupling dataset was constructed to provide a data basis for subsequent research. Based on this, this study designed an adaptive two-stage light coupling module LCM (Light Coupling Module), which significantly improved the robot’s perception data signal-to-noise ratio (SNR) and feature extraction robustness in dynamic lighting environment. Furthermore, combined with the adaptive two-stage light coupling module, an adaptive light enhancement model LEM (Light Enhancement Model) was proposed, and the Gaussian enhancement method GEM (Gausterity enhancement method) was innovatively introduced to expand the receptive field by fusing high-dimensional information and enhance the robustness of the model in dynamic lighting environment. In addition, this study also developed a robotic arm-vision system calibration method that incorporates a light compensation mechanism (LDM) to mitigate the impact of light on system performance, reducing the harvesting damage rate from 26.9% to 12.5%. The method was then integrated into an agricultural robot platform for verification. Experimental results show that compared to the original target detection model, the light coupling module (LCM) and LEM model designed in this study increased mAP (mean Average Precision) from 92.36% to 96.11%. Although the number of model parameters has increased, the detection accuracy and robustness in complex lighting environments have been significantly improved. In the actual operation test of the robot, after the introduction of the light coupling module LCM, LEM model and light compensation mechanism LDM, the system positioning accuracy reached 92.3%, which significantly improved the recognition accuracy and positioning accuracy, and greatly improved the operation efficiency of the agricultural robot. This study provides an innovative solution to solve the perception and operation problems of agricultural robots in complex lighting environments, and lays an important foundation for the implementation of smart farms and the development of modern agricultural robot technology.

Feng Dai was removed from the list of authors in the original publication