A5100772698- Advanced Adaptive Filtering Techniques
- Speech and Audio Processing
- Blind Source Separation Techniques
- Advanced Algorithms and Applications
- Data Mining Algorithms and Applications
- Urban and Freight Transport Logistics
- Metallurgy and Material Forming
- 3D IC and TSV technologies
- Advanced Computational Techniques and Applications
- Advanced Control Systems Design
- Face and Expression Recognition
- Control Systems and Identification
- Machine Learning and Data Classification
- Induction Heating and Inverter Technology
- VLSI and FPGA Design Techniques
- Advanced Manufacturing and Logistics Optimization
- Advanced Optical Imaging Technologies
- Surface Roughness and Optical Measurements
- Atomic and Subatomic Physics Research
- Rough Sets and Fuzzy Logic
- Semiconductor Lasers and Optical Devices
- Industrial Technology and Control Systems
- Gene expression and cancer classification
- Advanced Measurement and Detection Methods
- Wireless Sensor Networks and IoT
Chengdu University of Information Technology
2011-2025
Sichuan University
2019-2025
Shanghai Fudan Microelectronics (China)
2024
Shanghai University
2024
Zhengzhou University
2013
In this paper, a proportional integral (PI) controller that optimized with the modified different evolutional (DE) algorithm is proposed for speed control of brushless direct-current (BLDC) motor. The parameters PI are tuned by DE which based on adaptive mutation factor, multivariable fitness function and starting rule algorithm. performances controller, conventional standard (PI-SDE controller) investigated compared in simulation. Also, other optimization study. simulation results...
The Euclidean direction search (EDS) algorithm can reduce the complexity by avoiding matrix inversion operation. However, it may fail to work in impulsive environments. To address this problem, a novel EDS based upon maximum correntropy criterion (EDS-MCC) is proposed, which provides computational savings and robustness for combating noise. Additionally, EDS-MCC analyzed obtain theoretical performance utilizing energy conservation argument (ECA) Taylor expansion method. Simulations are...
This paper proposes a high linearity digital scanning algorithm (HLD), which quantifies the additional luminous current generated during switching process of driving transistor, and generates linear weights by adjusting time amplitude subfield, eliminating low gray level caused row data zeroing operation fractional subfield in traditional pulse width modulation (PWM) algorithm. The results indicate that modulating Vcom, RMSE grayscale is increased from 3.8689 to 2.8826. In addition, this...
The increasing complexity of modern very-large-scale integration (VLSI) design highlights the significance Electronic Design Automation (EDA) technologies. Chip placement is a critical step in EDA workflow, which positions chip modules on canvas with goal optimizing performance, power, and area (PPA) metrics final designs. Recent advances have demonstrated great potential AI-based algorithms enhancing placement. However, due to lengthy workflow design, evaluations these often focus...
Data missing is a vitally important issue that influences the classification results in medical field. This paper proposes an improved support vector machine (SVM) imputation algorithm by using strategies of pre-imputation, multiple iteration and grid search (IG-SVMI). Based on experimental performance, nine UCI datasets two real are used to compare proposed with four existing algorithms (RFI, KNNI, CCMVI orthogonal coding SVMI). The considered into types originally containing value randomly...
Strip flatness and thickness are two very important quality control indicators in the steel strip hot rolling process. According to strong coupling relation between parameters changing process of production, a system combined particle swarm optimization (PSO), model reference adaptive decoupling is designed. The adjusts controller online by ability PSO. results simulation show that this method effective on decoupling, reduce influence change effectively. has good robustness improve strip.
Aiming at the problems existed in process of mining manipulate errors for railway transport, an improved DBSCAN algorithm is presented. This method according to monitoring data operation train, transportation driver errors. Algorithm dealing with boundary shared object cluster by average distance function object. As a result it efficaciously avoids over-segmentation and produces better clustering results. Experimental results show that this effective.
In the process of seafood sales, choice transportation route directly affects efficiency and sales cost. this paper, based on investigation main aquatic products markets in Haikou, a mathematical model path planning is built with goal minimizing cost, simulated annealing algorithm used to solve problem.