Research On The Realization Technology Of Genetic Algorithms Based On FPGA

As so far, the research on Genetic Algorithms is focused on software implementation, it is always restricted by computer system which is essentially serial at speed. Therefore, it has a great meaning for hardware implementation.After having profound understanding on three basic operators of the Genetic Algorithms, according to the characteristic of hardware implementation, idiographic operators of GAs and parameter are confirmed. Then according to the request of algorithm and characteristic of FPGA ,the module was divided and the circuit structure was designed, this module includes these functions as follows: initialing population,calculating fitness,reproduction,corssover,mutation and population storing. Then has analysed function,port joining,inside structure of every module ,etc. in detail. Using hardware description language to program for function implementation, After function simulation,synthesis,place and route,timing simulation and download, the design is implemented in the Spartan 3 serial XC3S400- 4PQ208 chips of Xilinx .All procedure of design is worked under the ISE 6.2 integrated environment. Aiming at idiographic testing function,the test result shows: the design has realized basic functions of Genetic Algorithms. It does not consume too many resources of the chip, its speed of operation has been realized improving a lot more than software, it has solved the problem that the speed and area restrict each other in FPGA design and provided a possibility for the real-time using of Genetic Algorithms.Finally, by analyzing the design plan, it proposed several possible solutions to the shortage of the design.