摘要
遗传算法具有天然的并行性。FPGA( Field programmable gate arrays)本质上的并行特性使其很适合用于实现并行的遗传算法。结合两者的并行特性 ,本文提出了一种基于 FPGA的并行遗传算法。选用了适合硬件实现的选择、交叉、变异算子 ,并将它们设计成流水线结构。整个设计采用了 XILINX公司的 XC2 V1 0 0 0型号的FPGA芯片。算法利用 VHDL语言来描述。实现后的测试表明 ,这种硬件遗传算法有效减少了运行时间 。
Genetic algorithms (GAs) have the inherently parallel characteristic. The capability for parallelization of f ield programmable gate arrays (FPGA) makes it be excellent candidate for impleme nting the parallel GA. Combining the parallel characteristic of them together, a FPGA-based parallel genetic algorithm is proposed. The selection, crossover an d mutation operators suitable for the hardware implement are designed as a pipel ining architecture. The hardware GA processor is implemented in XILINX FPGA XC2V 1000. The VHDL language is used to describe the whole algorithm. Experimental re sults indicate that the hardware-based parallel genetic algorithm can efficient ly reduce the run time. It can be applied in real-time system applications.
出处
《南京航空航天大学学报》
EI
CAS
CSCD
北大核心
2004年第5期628-632,共5页
Journal of Nanjing University of Aeronautics & Astronautics
基金
国防基础科研基金资助项目
关键词
遗传算法
并行
现场可编程门阵列
VHDL
genetic algorithms
parallel
field programmable gate arrays (FPGA)
VHDL
