Research on evolution hardware design based on memetic algorithm

When traditional evolution algorithms are used in hardware evolution, it has the disadvantages of slow evolution speed and premature convergence. In order to overcome the shortcomings, a memetic algorithm(MA) that combines the local search method with genetic algorithm is proposed for combinational logic circuit design based on CGP. A gate type local search strategy is proposed in order to be adaptive for hardware design based on FPGA. The experiments are carried out by designing 1-bit full subtractor. Experimental results show that it is much more efficient in solving the evolution hardware problem than GA and particle swarm optimization.