Research on Hardware Architecture for Particle Swarm Optimization Positioning System

• Main Authors: Kui-Ting Chen, 陳奎廷
• Other Authors: Kuo-Yang Tu
• Format: Others
• Language: en_US
• Published: 2009
• Online Access: http://ndltd.ncl.edu.tw/handle/43084159168831145642

碩士 === 國立高雄第一科技大學 === 系統資訊與控制研究所 === 97 === Nowadays, many researchers begin conducting research on positioning technologies which are applica¬ble to daily life. The positioning task becomes one of the popular re-search topics. Many algorithms have been proposed for positioning system. Compar¬ing with other evolutionary algorithms, such as Genetic Algorithm (GA), Particle Swarm Optimization (PSO) can achieve higher accuracy and faster convergence for the posi-tion¬ing issue, but the software implementation of PSO need much calculation time to estimate the position of target. The PSO software implementation only can achieve 25 times of target positioning in one second. In real-time environment, the software implementation approach is hardly to achieve high accuracy. However the hardware implementation of PSO has never been used for positioning system in previous research. This research focuses on the PSO implementation of hardware combined with software for the real-time positioning systems. In order to reduce the calculation time, the SOPC-based hardware architecture of PSO is proposed in this study. We selected PSO Random Time-varying Inertial Weight and Accelerate Coefficient (PSO-RTVIWAC) algorithm to deal with optimization tasks in positioning system. The experimental results show that the SOPC-based architecture is faster enough to deal with real-time positioning issue. Moreover, SOPC-based architecture can estimate position of target with a speed of 100 times per second.