A5102750992- Rock Mechanics and Modeling
- Landslides and related hazards
- 3D Surveying and Cultural Heritage
- Geotechnical Engineering and Analysis
- Geotechnical Engineering and Soil Mechanics
- 3D Shape Modeling and Analysis
- Geotechnical Engineering and Underground Structures
- Drilling and Well Engineering
- Hydrocarbon exploration and reservoir analysis
- Autonomous Vehicle Technology and Safety
- Robotics and Sensor-Based Localization
- Robotic Path Planning Algorithms
- Paleontology and Stratigraphy of Fossils
- Natural Fiber Reinforced Composites
- Geophysical Methods and Applications
- Grouting, Rheology, and Soil Mechanics
- Anomaly Detection Techniques and Applications
- Vehicle Dynamics and Control Systems
- Advanced Image and Video Retrieval Techniques
- Magnetic Bearings and Levitation Dynamics
- Karst Systems and Hydrogeology
- Machine Learning and ELM
- Odor and Emission Control Technologies
- Polymer Nanocomposite Synthesis and Irradiation
- Energy Load and Power Forecasting
Chengdu University of Technology
2020-2025
Chongqing University of Posts and Telecommunications
2023-2024
Zhejiang University
2023
University of Alicante
2022
Hebei University of Science and Technology
2021
Chang'an University
2019
Chung Shan Medical University
2001
For years, the YOLO series has been de facto industry-level standard for efficient object detection. The community prospered overwhelmingly to enrich its use in a multitude of hardware platforms and abundant scenarios. In this technical report, we strive push limits next level, stepping forward with an unwavering mindset industry application. Considering diverse requirements speed accuracy real environment, extensively examine up-to-date detection advancements either from or academia....
Density-based and classification-based methods have ruled unsupervised anomaly detection in recent years, while reconstruction-based are rarely mentioned for the poor reconstruction ability low performance. However, latter requires no costly extra training samples that is more practical, so this paper focuses on improving kind of method proposes a novel Omni-frequency Channel-selection Reconstruction (OCR-GAN) network to handle task perspective frequency. Concretely, we propose Frequency...
As the proportion of wind power in world's electricity generation increases, improving prediction accuracy is vital for making full use energy and ensuring safe stable operation grid. Given uncertainty volatility weak generalization ability current models, we propose a model that combines Adaboost algorithm with extreme learning machine optimized by particle swarm optimization (PSO-ELM). First, used to optimize initial thresholds input weights ELM obtain PSO-ELM basic model. Then, combined...
ABSTRACT Investigating the creep properties of damaged rocks is essential for evaluating long‐term stability seismically cracked slopes in earthquake‐prone regions. In this study, cyclic loading‐unloading and step loading tests were sequentially performed on metamorphic sandstone, granite, phyllite. Dissipated energy was introduced to establish initial damage model, effect patterns mechanisms analyzed. The experimental results showed that dissipated variable increased linearly with number...
This paper develops a flexible collision avoidance strategy for the formation of multiple unmanned aerial vehicles (multi-UAV). Firstly, To make full use information flow from communication formation, an improved artificial potential field (IAPF) function is constructed by adding topology and weights. Based on IAPF functions, both within obstacle outside are well organized to ensure flight safety. Secondly, fuse keeping, we design null space behavioral (NSB) approach combine them into single...
Pairwise point cloud registration is a critical task for many applications, which heavily depends on finding correct correspondences from the two clouds. However, low overlap between input clouds causes to fail easily, leading mistaken overlapping and mismatched correspondences, especially in scenes where non-overlapping regions contain similar structures. In this paper, we present unified bird's-eye view (BEV) model jointly learning of 3D local features estimation fulfill pairwise loop...
Acoustic emission and thermal infrared signals are related to the process of rock breaking. In this paper, acoustic obtained during loading indoor granite freeze-thaw cycle test uni-axial compression test. Then sent out breaking analyzed by short-time Fourier transform get peak frequency. radiation thermographic sensor is applied monitor record temperature change on surface so as influence characteristics materials deformation Under same axial stress, samples passing through more cycles have...
Current automatic parking systems often separate path planning and trajectory tracking, increasing complexity failing to meet the needs of autonomous cars. Consequently, this study presents an approach based on Multi-stage Nonlinear Model Predictive Control (MSNMPC) for integrated tracking optimization. The method represents evolution uncertainty parameters over time through a scenario tree, MSNMPC controller with predictive horizon P defines + 1 phases, each specific cost constraint...
The creep behaviors of prepared PVA-ECC materials(fiber content 1.5%) and normal concrete were tested through stress test at a level 35%, along with shrinkage during early age(0-28d). Results indicate that material exhibits lower coefficient 1.45 times higher than the same strength age.
Daguangbao (DGB) Landslide (12 × 108 m3) is the largest landslide triggered by 2008 Ms8.0 Wenchuan Earthquake, in which basal shear failure develops on an interlayer fault at 400 m deep under ground. After landslide, a 1.8 km long (in sliding direction) oblique face exposed. Different kinds of materials DGB are taken for direct test, medium scale in-situ test and ring test. The results show that material cohesion ranges from 20 to 320 kPa internal friction angle 15° 41°. Shearing strength...
Pairwise point cloud registration is a critical task for many applications, which heavily depends on finding correct correspondences from the two clouds. However, low overlap between input clouds causes to fail easily, leading mistaken overlapping and mismatched correspondences, especially in scenes where non-overlapping regions contain similar structures. In this paper, we present unified bird's-eye view (BEV) model jointly learning of 3D local features estimation fulfill pairwise loop...
In the complex and changeable driving environment, loader's autonomous decision-making planning ability is seriously insufficient, it cannot make predictions about upcoming behaviors based on historical experience current environment like human drivers. view of this, we carry out following research. The research background typical operation mode loader with extremely frequent braking action, core prediction brake pedal aperture for automatic wheel loader. By combining data experienced...
During the development of urban underground space in Chengdu, China, it was found that glauberite rocks Guankou formation Cretaceous system had extensive dissolution and exerted a serious impact on safety project. At present, research transport mechanical properties under hydrostatic conditions this area is limited. In paper, effects seepage are studied. The results show concentration aqueous solution increases step by with increase time still water, indicating forward movement front rock...