A5062605255- Adsorption and Cooling Systems
- Solar-Powered Water Purification Methods
- Heat Transfer and Optimization
- Refrigeration and Air Conditioning Technologies
- Building Energy and Comfort Optimization
- Chaos-based Image/Signal Encryption
- Cellular Automata and Applications
- Traffic Prediction and Management Techniques
- Transportation Planning and Optimization
- Adaptive Control of Nonlinear Systems
- Robotic Mechanisms and Dynamics
- Advanced Steganography and Watermarking Techniques
- Fractal and DNA sequence analysis
- Greenhouse Technology and Climate Control
- Video Surveillance and Tracking Methods
- Traffic control and management
- Advanced Neural Network Applications
- Solar Thermal and Photovoltaic Systems
- Iterative Learning Control Systems
- Industrial Vision Systems and Defect Detection
Xinjiang University
2024
Shanghai Jiao Tong University
2017-2019
Abstract Comfortable, efficient, and affordable heating, ventilation, air conditioning systems in buildings are highly desirable due to the demands of energy efficiency environmental friendliness. Traditional vapor-compression conditioners exhibit a lower coefficient performance (COP) (typically 2.8–3.8) owing cooling-based dehumidification methods that handle both sensible latent loads together. Temperature- humidity-independent control or desiccant have been proposed overcome these...
Aiming at the problems of target detection models in traffic scenarios including a large number parameters, heavy computational burden, and high application cost, this paper introduces an enhanced lightweight real-time algorithm, which exhibits higher speed accuracy for vehicle detection. This considers YOLOv7 algorithm as benchmark model, designs backbone network, uses MobileNetV3 network to extract features. Inspired by structure SPPF, spatial pyramid pooling module is reconfigured...
Accurately predicting traffic flow is crucial for optimizing conditions, reducing congestion, and improving travel efficiency. To explore spatiotemporal characteristics of in depth, this study proposes the MFSTBiSGAT model. The model leverages graph attention networks to extract dynamic spatial features from complex road networks, utilizes bidirectional long short-term memory capture temporal correlations both past future time perspectives. Additionally, information enhancement layers are...
This paper presents a digital image encryption scheme based on an amnesic super chaotic system. First, the performance of hyperchaotic system with addition memristor is analyzed, such as phase track map, Lyapunov exponential Pongalai cross-section, and other tests. The test results prove that has better hyper-chaotic properties. proposed in this generates initial value through SHA3-256 function to realize 'one one secret.' Chaotic sequences are generated by memristive used encrypt image....
Abstract To enhance the precision of trajectory tracking control in industrial robots, a nonlinear end-to-end sliding mode algorithm for robots based on dung beetle is proposed. By employing terminal to resolve issue inadequate performance and integrating optimizer (DBO) parameter adjustment, significant enhancements are achieved accuracy.
Abstract A layer multiplexing-based encrypted image scheme for DNA color images is proposed. Firstly, a memristor added to the Sprott-B chaotic system make it fourth-order memristorised hyper-chaotic system. The hyperchaotic has better pseudo-randomness. encryption proposed generates initial value of by SHA3-256 function achieve “One image, one secret”. sequence generated memory encrypting image. process includes zero complements, chunking, encryption, multiplexing, decryption, and improved...