This study aims to address the current situation of the late start of mechanized harvesting technology for the pod pepper, the high damage rate of existing pod pepper harvesters, and the lack of theoretical support for key harvesting components. The Hertz theory is employed to investigate the damage mechanism of collisions between pod pepper and comb fingers. The study analyzes the maximum deformation of pod pepper and the critical speed at which damage occurs during the collision process. Furthermore, it explores the critical relative speed that leads to damage in pod pepper. Orthogonal tests are conducted to analyze the effects of rotational speed, hose thickness, and moisture content on the efficiency of pod pepper picking. The experimental results are then subjected to multifactorial ANOVA to identify the optimal test parameters. The structural and motion parameters of the picking device are optimized based on these conditions. It is determined that the critical relative velocity for damage to pod pepper during a collision with the comb finger is V0 = 11.487 m s−1. The collision velocities of pod pepper with different hose thicknesses are analyzed using the i-SPEED TR endoscopic high-speed dynamic analysis system to obtain the corresponding collision velocities for different hose thicknesses. The study finds that rotational speed, hose thickness, and the water content of pod pepper affect the damage rate and stem shedding rate. The optimal experimental parameters are determined to be a rotational speed of 705.04 rpm, hose thickness of 3 mm, and water content of the pepper of 71.27%.
7 July 2025 in IEEE Transactions on Vehicular Technology
As vehicular network services evolve, the limitations of relying solely on terrestrial networks for communication become increasingly apparent, particularly in terms of network coverage and reliability. Integrating software defined network (SDN) and network function virtualization (NFV) with satellite networks can significantly enhance their networking capabilities. This integration facilitates the formation of satellite-terrestrial integrated vehicular networks (STIVNs), which provide extensive and reliable communication services. A key component in this context is the service function chain (SFC), a prevalent SDN/NFV network service paradigm. The deployment of virtual network functions (VNFs) within an SFC is crucial for optimizing network performance. In this paper, we introduce a novel three-layer architecture for STIVNs, termed layered virtual service satellite-terrestrial integrated vehicular networks (LVS-STIVNs). To optimize quality of service (QoS) and quality of experience (QoE) from the user perspective, we model the dynamic deployment of SFCs as a Markov decision process (MDP). To further enhance adaptability to the dynamic nature of services in satellite networks, we propose an orchestration method based on deep reinforcement learning (DRL), named rDQN-SFC. This method leverages long-short-term memory (LSTM) to effectively capture sequential states. The results indicate that the QoE performance achieved by the rDQN-SFC algorithm is at least 2.02 times that of other DRL baseline algorithms, and the acceptance rate of dynamic SFC deployment is improved by at least 51.1%, demonstrating a more stable adaptive deployment capability in dynamic environments. Additionally, the algorithm exhibits a high time-efficiency ratio, delivering superior service performance in a shorter duration.
3 August 2018 in IEEE Transactions on Circuits and Systems II: Express Briefs
Dictionary learning for sparse tensor representation aims to train a set of dictionaries for each dimension using tensor samples based on the Tucker decomposition. However, their applications are limited by the fact that all the training samples must be input simultaneously, and that there is no direct way to extend the existing online dictionary learning methods for the vector-based model to the Tucker-decomposition-based model. To overcome this limitation, in this study, we develop a Tucker-decomposition-based strategy to achieve a warm start for updating dictionaries, based on which, an online tensor dictionary learning algorithm is proposed. The proposed algorithm processes a single new training sample at a time, such that it can be used not only for offline learning from static samples, but also for online learning from dynamic samples, under the framework of the Tucker model. When new training samples are input, only the newly-added samples need to be used for retraining the dictionaries. We verify the convergence and low-complexity of the proposed algorithm via theoretical analysis as well as online learning simulations. We also perform offline learning simulations, the results of which demonstrate that our algorithm has an obvious advantage in training accuracy over existing tensor dictionary learning algorithms. The proposed algorithm has the potential to be used in fields such as multi-dimensional signal processing, compressive sensing, and machine learning.
10 November 2025 in IEEE Journal of Selected Topics in Signal Processing
Spatiotemporal fusion (STF) enables the integration of remote sensing (RS) images with high spatial and temporal resolutions, playing a vital role in agricultural applications such as crop monitoring and disease/pest early warning. Recently, Transformer has been applied in the field of STF and has aroused widespread attention due to its powerful representation ability for capturing global long-term dependence features. However, the Transformer applied to STF utilizes self-attention with spatial windows, coupled with channel-wise weight sharing, which significantly constrains cross-spectral adaptive modeling. In addition, conventional feature-extraction frameworks exhibit substantial attenuation of high-frequency components, resulting in spatially blurred fusion outputs. In this study, we propose a Dual-Interaction Network (DINet) for STF of agricultural RS. The DINet operates on two cross-paired inputs (a fine reference image and a coarse target image), significantly relaxing the strict requirements of conventional STF methods. In the proposed framework, a spatial-channel Transformer (SCT) is developed to adjust receptive fields in light of the distinct resolutions of coarse/fine inputs, enhancing spatial-channel feature interaction during the feature-extraction stage. A wavelet decomposition-enhanced strategy is leveraged to integrate high-frequency details and low-frequency global information through a spatial-temporal Transformer (STT). In the reconstruction stage, a multi-level fusion module (MLFM) is designed to hierarchically fuse the features from multiple branches, boosting the spatial-temporal consistency. Experimental results demonstrate that the proposed method outperforms state-of-the-art approaches in preserving fine-grained agricultural features while maintaining temporal consistency, achieving superior performance for crop monitoring tasks.
Control accuracy significantly affects the performances of boom sprayer. In this study, we develop a precise autocontrol technology based on the vehicle speed feedback. We utilize the auxiliary antidrift system of wind-curtain type air flow and the variable spraying control system for adaptive fertilizing and online measuring of working conditions. Experimental results demonstrate that the variable spraying control system could keep the speed error less than 3%. The air flow significantly improves the penetration of spraying, decreases the fog drip, and increases the pesticide utility. Benefitting from the auxiliary air flow, the average utility of pesticide is improved from 26.76% to 37.98%. Additionally, the speed feedback control reduces the consumption of pesticide by more than 12%.
23 February 2026 in IEEE Transactions on Industrial Informatics
In industrial scenarios, data augmentation is an effective approach to improve model performance. However, its benefits are not unidirectionally beneficial. There is no theoretical research or established estimation for the optimal sample size (OSS) in augmentation, nor is there an established metric to evaluate the accuracy of OSS or its deviation from the ground truth. To address these issues, we propose an information-theoretic optimal sample size estimation (IT-OSE) to provide reliable OSS estimation for industrial data augmentation. An interval coverage and deviation (ICD) score is proposed to evaluate the estimated OSS intuitively. The relationship between OSS and dominant factors is theoretically analyzed and formulated, thereby enhancing the interpretability. Experiments show that, compared to empirical estimation, the IT-OSE increases accuracy in classification tasks across baseline models by an average of 4.38%, and reduces mean absolute percentage error (MAPE) in regression tasks across baseline models by an average of 18.80% . The improvements in downstream model performance are more stable. $\mathbf {ICD_{\mathbf {\text{dev}}}}$ in the ICD score is also reduced by an average of 49.30% . The determinism of OSS is enhanced. Compared to exhaustive search, the IT-OSE achieves the same OSS while reducing computational and data costs by an average of 83.97% and 93.46% . Furthermore, practicality experiments demonstrate that the IT-OSE exhibits generality across representative sensorbased industrial scenarios.