We develop an accelerated Genetic Algorithm (GA) system constructed by the cooperation of field-programmable gate array (FPGA) and optimized parameters of the GA. We found the enhanced decay of mutation rate makes convergence of the GA much faster, enabling the parameter-induced acceleration of the GA. Furthermore, the accelerated configuration of the GA is programmed in FPGA to boost processing speed at the hardware level without external computation devices. This system has ability to focus light through scattering medium within 4 seconds with robust noise resistance and stable repetition performance, which could be further reduced to millisecond level with advanced board configuration. This study solves the long-term limitation of the GA, it promotes the applications of the GA in dynamic scattering mediums, with the capability to tackle wavefront shaping in biological material.
翻译:我们通过现场可编程门阵列(FPGA)和最佳大会参数的合作,开发了一个加速遗传算法(GA)系统。我们发现突变率的加速衰减使大会的趋同速度加快得多,使大会的参数引致加速。此外,GA加速配置在FPGA中编程,在没有外部计算装置的情况下提高硬件级的处理速度。这个系统有能力在4秒钟内通过散射介质来聚焦光线,且具有很强的抗噪声阻力和稳定的重复性能,这可以进一步降低到毫秒水平,并配有高级板配置。本研究解决了GA的长期限制,它促进了GA在动态散射介质中的应用,能够处理生物材料的波形成。