A5114734444- Urban Stormwater Management Solutions
- Flood Risk Assessment and Management
- Cellular Automata and Applications
- Hydrology and Watershed Management Studies
- Urban Green Space and Health
- Advanced Data Storage Technologies
- Remote Sensing and Land Use
- Land Use and Ecosystem Services
- Handwritten Text Recognition Techniques
- Phase-change materials and chalcogenides
- Evaluation Methods in Various Fields
- Seismic Waves and Analysis
- Hand Gesture Recognition Systems
- Seismic Imaging and Inversion Techniques
- Urban Heat Island Mitigation
- Adaptive Dynamic Programming Control
- Reinforcement Learning in Robotics
- Regional Economic and Spatial Analysis
- Seismology and Earthquake Studies
- Semiconductor Lasers and Optical Devices
- Air Quality Monitoring and Forecasting
- Distributed Control Multi-Agent Systems
Tsinghua University
2003-2024
Seismic noise suppression is an important preprocessing stage for obtaining high-quality seismic signals, which are crucial exploration. Deep learning methods have achieved excellent results in the field of signal processing. Currently, many researchers used convolutional neural networks denoising, but few Transformer model related research. We apply swin transformer model, improved version based on self-attention mechanism, to denoise two-dimensional data. The calculates within shifted...
Coping with environmental and climate changes induced by rapid urbanization has become a major challenge for urban blue-green spatial planning management. This study explores the effects of green belts, wedges, neighborhood structures on carbon sinks to help understand multi-scale structure-function response patterns. The results show that density patches within belts wedges was lower than built-up areas. near farmland grassland water bodies forests. Additionally, forest-water structure is...
Handwritten and machine-printed characters are recognized separately in most OCR systems due to their distinct difference. In applications where both kinds of involved, it is necessary judge a character's handwritten/printed property before feeding into the proper recognition engine. this paper, new method discriminate between handwritten character proposed. Unlike previous works, discrimination we carried out paper totally based on single character. Five statistical features extracted from...
Disc scratches, dusts and noise corruption et al. will cause data sequence errors including random burst errors. With the recording density of multilevel (ML) optical storage system increasing, same length scratches on surface disc bring more in ML DVD than conventional systems, e.g., digital versatile (DVD) compact (CD). A new interleaving scheme based Reed-Solomon code (RS) is proposed to solve this problem. In system, burden Parity Outer-code (PO) huge, while Innercode (PI) has correcting...
Multilevel modulation coding is the core technology for multilevel optical storage. The employed to increase capacity of read-only disc with no change on original recording symbols size. With combination and run-length limited modulation, short low levels can form long high while it also suitable by run-lengths. And DC-suppression be controlled allowance without sacrifice effective bits. This double-modulation scheme not only improve efficiency, but greatly density. In this paper, theory...
A new modulation coding method for signal waveform multilevel disc is presented. The proposed code uses run-length and level to satisfy the channel constraints. density ratio information bits per 400 nm are 1.67 2.08 respectively.
Climate change-induced storms have posed a threat to current passive urban stormwater systems. As critical flood control measure, static blue-green infrastructure (BGI) is incapable of effectively utilizing available storage capacity mitigate peak flow during cloudburst events. At storm times, landscape-based BGI may overflow the drainage system, escalating downstream flooding risks. To alleviate effect, we investigated an adaptive RTC solution that would predischarge and initiate runoff...