scholarly journals Design of High-efficiency LED Driver Power Supply Based on Improved Particle Swarm Optimization

2021 ◽  
Vol 2136 (1) ◽  
pp. 012021
Author(s):  
Fanghui Zhang ◽  
Huimin Qiu
Keyword(s):  
Particle Swarm Optimization ◽  
Particle Swarm ◽  
Soft Switching ◽  
Constant Current ◽  
Resonant Converter ◽  
Led Driver ◽  
Swarm Optimization ◽  
Improved Particle Swarm Optimization ◽  
Photoelectric Properties ◽  
Llc Resonant Converter

Abstract The photoelectric properties of LED are affected by driving current and ambient temperature. To solve the problem of light wane caused by excessive current and overheating, LLC resonant converter with soft switching was chosen as the main topology, improving the power efficiency, and reducing the heat of the LED driver. Firstly, the soft-switching conditions of the LLC resonant converter are analyzed, and the constraint equation is obtained. Then, an improved particle swarm optimization algorithm is proposed to optimize the resonant parameters. The weight of particles is dynamically and adaptively changed to balance the global and local optimization performance of particles. Finally, the controller is designed to realize constant current. This design improves the efficiency of the LED driver. At the same time, the light wane is effectively suppressed by stable output current, and the rationality of the selected parameters is verified by simulation and experiment.

Compressed Sensing Image Reconstruction Based on Improved Particle Swarm Optimization Algorithm

2014 ◽  
Vol 599-601 ◽  
pp. 1453-1456
Author(s):  
Ju Wang ◽  
Yin Liu ◽  
Wei Juan Zhang ◽  
Kun Li
Keyword(s):  
Particle Swarm Optimization ◽  
Image Reconstruction ◽  
Compressed Sensing ◽  
Optimization Algorithm ◽  
Particle Swarm Optimization Algorithm ◽  
Matching Pursuit ◽  
Particle Swarm ◽  
Premature Convergence ◽  
Swarm Optimization ◽  
Improved Particle Swarm Optimization

The reconstruction algorithm has a hot research in compressed sensing. Matching pursuit algorithm has a huge computational task, when particle swarm optimization has been put forth to find the best atom, but it due to the easy convergence to local minima, so the paper proposed a algorithm ,which based on improved particle swarm optimization. The algorithm referred above combines K-mean and particle swarm optimization algorithm. The algorithm not only effectively prevents the premature convergence, but also improves the K-mean’s local. These findings indicated that the algorithm overcomes premature convergence of particle swarm optimization, and improves the quality of image reconstruction.

Sign in / Sign up

Export Citation Format

Share Document