Realization of swarm intelligence by spiking neurons, and its new development

在这项研究中，我们开发了利用群体智能优化各种问题的非线性程序，并考虑了神经元放电的硬件实现及其在工程上的应用。因此，我们实现了粒子受确定性动力学系统影响的群体智能优化技术。该系统基于离散时间和异步系统的神经元模型，利用混沌动力学的搜索机制，成功地实现了系统的高性能优化技术。在此基础上，实现了整型确定性粒子群算法，便于在现场可编程门阵列上实现，并给出了算法的运行速度。

In this study, we developed nonlinear procedure to optimize various problems by swarm intelligence, and the hardware realization by spiking neurons and the application for the engineering are considered. Therefore we realized the swarm intelligence optimization technique with a particle influenced by deterministic dynamics system with spiking neurons. The system is based on a neuronal model of the discrete time and asynchronous system, and by applying the searching mechanism that used chaos dynamics, we succeeded in realization of high-performance optimization technique in the system. Furthermore, we realized deterministic particle swarm optimization procedure of the integer type to facilitate implementation in FPGA and showed what it operates fast.