Zhenqiao Hui
Zhenqiao Hui has a total of 7 co-authors. Here are the top co-authors by number of shared publications:
- Jichao Wang
Hebei University of Water Resources and Electric Engineering
2 shared publications - Donglai Sun
Hebei University of Water Resources and Electric Engineering
1 shared publication - Hanyu Zhang
Hebei University of Water Resources and Electric Engineering, China
1 shared publication - Mengyuan Li
Hebei University of Water Resources and Electric Engineering
1 shared publication - Shuai Zhang
Hebei University of Water Resources and Electric Engineering, China
1 shared publication Sicong Pang
Hebei University of Water Resources and Electric Engineering
1 shared publication
Latest Publications
- Open Access
- 8 December 2025
- Published by MDPI AG in Preprints.org
ED-MOEA: An Event-Driven Multi-Objective Evolutionary Algorithm for Cooperative Planning of Heterogeneous UAV-UGV Systems in Dynamic Post-Disaster Environments
Post-disaster rescue operations require rapid and efficient coordination of heterogeneous 2 unmanned systems to locate survivors and deliver supplies in dynamic, uncertain environ- 3 ments. Existing approaches predominantly adopt a static task allocation paradigm, which 4 fails to adapt to scenarios where new survivors are continually discovered and obstacles 5 appear unexpectedly. This paper proposes ED-MOEA (Event-Driven Multi-Objective Evolu- 6 tionary Algorithm), establishing a closed-loop cooperative sensing-planning mechanism for 7 heterogeneous UAV-UGV systems. The core innovation lies in an event-triggered dynamic 8 replanning strategy: when UAVs detect new task points or obstacles during flight, the 9 system immediately broadcasts information to UGVs and triggers a warm-start replanning 10 process. The problem is formulated as a three-objective optimization considering makespan 11 minimization, energy consumption minimization, and task coverage maximization, while 12 accounting for heterogeneous mobility constraints—UAVs can fly directly over obstacles 13 while UGVs must navigate around them. Experiments on five benchmark scenarios and 14 a real case based on the 2024 Hualien earthquake demonstrate that ED-MOEA achieves 15 the highest hypervolume (HV) across all scenarios, outperforming existing multi-objective 16 algorithms including NSGA-III, MOEA/D-DE, RVEA, and LMEA. The heterogeneous 17 coordination achieves 16.7% higher task coverage compared to UAV-only systems (0.910 18 vs. 0.780), while the warm-start strategy significantly accelerates replanning convergence 19 (over 35% reduction in generations required), ensuring real-time response capability under 20 6 seconds for scenarios with up to 70 tasks.
- 8 December 2025
- Published by MDPI AG in Preprints.org
Open AccessED-MOEA: An Event-Driven Multi-Objective Evolutionary Algorithm for Cooperative Planning of Heterogeneous UAV-UGV Systems in Dynamic Post-Disaster Environments
Post-disaster rescue operations require rapid and efficient coordination of heterogeneous 2 unmanned systems to locate survivors and deliver supplies in dynamic, uncertain environ- 3 ments. Existing approaches predominantly adopt a static task allocation paradigm, which 4 fails to adapt to scenarios where new survivors are continually discovered and obstacles 5 appear unexpectedly. This paper proposes ED-MOEA (Event-Driven Multi-Objective Evolu- 6 tionary Algorithm), establishing a closed-loop cooperative sensing-planning mechanism for 7 heterogeneous UAV-UGV systems. The core innovation lies in an event-triggered dynamic 8 replanning strategy: when UAVs detect new task points or obstacles during flight, the 9 system immediately broadcasts information to UGVs and triggers a warm-start replanning 10 process. The problem is formulated as a three-objective optimization considering makespan 11 minimization, energy consumption minimization, and task coverage maximization, while 12 accounting for heterogeneous mobility constraints—UAVs can fly directly over obstacles 13 while UGVs must navigate around them. Experiments on five benchmark scenarios and 14 a real case based on the 2024 Hualien earthquake demonstrate that ED-MOEA achieves 15 the highest hypervolume (HV) across all scenarios, outperforming existing multi-objective 16 algorithms including NSGA-III, MOEA/D-DE, RVEA, and LMEA. The heterogeneous 17 coordination achieves 16.7% higher task coverage compared to UAV-only systems (0.910 18 vs. 0.780), while the warm-start strategy significantly accelerates replanning convergence 19 (over 35% reduction in generations required), ensuring real-time response capability under 20 6 seconds for scenarios with up to 70 tasks.
- Open Access
- 11 November 2025
- Published by Springer Nature in Research Square
SF-YOLO11: A Real-Time Winter Jujube Detection Model Based on Lightweight Multi-Scale Fusion
To address the challenges of detecting winter jujubes in orchard environments—including dense small targets, variable illumination, complex backgrounds, and limited edge deployment capabilities—this paper proposes SF-YOLO11, a lightweight real-time detection model. First, a ternary channel importance metric is proposed that integrates gradient, variance, and task relevance to construct an adaptive pruning strategy. This strategy optimizes C3K2 into PrunedC3K2, achieving 31% model compression. Second, the LightSPPF module is designed through computational graph reconstruction. This module employs a three-stage architecture comprising depthwise separable convolution, adaptive pooling, and dynamic fusion, reducing computational cost by 40%. Finally, we transform multi-scale feature processing into sequence modeling and design the ISSFF module to capture scale dependencies through bidirectional LSTM and multi-head attention mechanisms, thereby improving small target detection accuracy. Experimental results demonstrate that SF-YOLO11 achieves an mAP@0.5 of 92.89%, representing a 3.68% improvement over the baseline. The model contains 1.6M parameters, requires 4.3 GFLOPs of computation, and operates at 128 FPS, achieving optimizations of 38.5%, 33.8%, and 14%, respectively. After INT8 quantization, the model size is reduced to only 1.7 MB, and the FPS increases to 215. In cross-fruit transfer experiments, the model achieves zero-shot mAP@0.5 values of 68.34% and 71.52% on grape and cherry tomato datasets, respectively. After fine-tuning, these values improve to 85.23% and 87.12%, respectively. Robustness tests validate the model's performance under extreme conditions, including strong illumination, low illumination, and occlusion.
- 11 November 2025
- Published by Springer Nature in Research Square
Open AccessSF-YOLO11: A Real-Time Winter Jujube Detection Model Based on Lightweight Multi-Scale Fusion
To address the challenges of detecting winter jujubes in orchard environments—including dense small targets, variable illumination, complex backgrounds, and limited edge deployment capabilities—this paper proposes SF-YOLO11, a lightweight real-time detection model. First, a ternary channel importance metric is proposed that integrates gradient, variance, and task relevance to construct an adaptive pruning strategy. This strategy optimizes C3K2 into PrunedC3K2, achieving 31% model compression. Second, the LightSPPF module is designed through computational graph reconstruction. This module employs a three-stage architecture comprising depthwise separable convolution, adaptive pooling, and dynamic fusion, reducing computational cost by 40%. Finally, we transform multi-scale feature processing into sequence modeling and design the ISSFF module to capture scale dependencies through bidirectional LSTM and multi-head attention mechanisms, thereby improving small target detection accuracy. Experimental results demonstrate that SF-YOLO11 achieves an mAP@0.5 of 92.89%, representing a 3.68% improvement over the baseline. The model contains 1.6M parameters, requires 4.3 GFLOPs of computation, and operates at 128 FPS, achieving optimizations of 38.5%, 33.8%, and 14%, respectively. After INT8 quantization, the model size is reduced to only 1.7 MB, and the FPS increases to 215. In cross-fruit transfer experiments, the model achieves zero-shot mAP@0.5 values of 68.34% and 71.52% on grape and cherry tomato datasets, respectively. After fine-tuning, these values improve to 85.23% and 87.12%, respectively. Robustness tests validate the model's performance under extreme conditions, including strong illumination, low illumination, and occlusion.