基于硬件实现的基因算法的研究

本文提出了一种VLSI实现的硬件基因算法.研究了基因算法的各种变种,探讨了它们的性能及硬件实现的可能性.提出了一个能进行群体存储、父本选择、交叉、变异等操作且易于硬件实现的结构.在硬件实现上,用VHDL描述了整个算法.所作的设计是一个通用的VLSI结构,通过流水线结构和并行化操作获得了很好的性能.硬件实现基因算法有效地缩短运行时间,为实时应用提供了可能.整个设计用Altera公司的FLEX10K40型号的芯片进行了FPGA实现,它完全可以用VLSI来实现.

Research on Hardware-based Genetic Algorithms

A VLSI implementation of hardware genetic algorithm is proposed in this paper.All kinds of variations of genetic algorithms are investigated,and their performance and the possibility of hardware implementation are discussed.An architecture that can perform the functions of population storage,selection,crossover,mutation,fitness evaluation,and survival determination,and is advantageous to hardware implementation is designed.When implementing the design in hardware,the VHDL language is used to describe the whole algorithms.The whole design is a general purpose VLSI architecture for GA.It achieves better performance by pipeline and parallelism.Hardware implementation of GA can reduce the computation time drastically and makes it possible to apply in real time applications.The Hardware GA processor has been implemented in Altera FPGA FLEX10K40.It can be realized in VLSI easily.