王秀青, 曾慧, 韩东梅, 刘颖, 吕锋. 基于脉冲神经网络的类脑计算[J]. 北京工业大学学报, 2019, 45(12): 1277-1286. DOI: 10.11936/bjutxb2018100018
    引用本文: 王秀青, 曾慧, 韩东梅, 刘颖, 吕锋. 基于脉冲神经网络的类脑计算[J]. 北京工业大学学报, 2019, 45(12): 1277-1286. DOI: 10.11936/bjutxb2018100018
    WANG Xiuqing, ZENG Hui, HAN Dongmei, LIU Ying, LÜ Feng. Brain-inspired Computing Based on Pulsed Neural Networks[J]. Journal of Beijing University of Technology, 2019, 45(12): 1277-1286. DOI: 10.11936/bjutxb2018100018
    Citation: WANG Xiuqing, ZENG Hui, HAN Dongmei, LIU Ying, LÜ Feng. Brain-inspired Computing Based on Pulsed Neural Networks[J]. Journal of Beijing University of Technology, 2019, 45(12): 1277-1286. DOI: 10.11936/bjutxb2018100018

    基于脉冲神经网络的类脑计算

    Brain-inspired Computing Based on Pulsed Neural Networks

    • 摘要: 针对当前重要国际科技前沿——类脑计算进行研究,讨论了类脑计算的研究内容、特点和研究现状.类脑计算不是简单的人脑神经元的模拟和神经元模型的应用,而是对人脑的信息处理规律、复杂的工作模式及思维、学习、推理、决策本质性机理的深层次模拟.脉冲神经网络比传统神经网络具有更好的生物似真性,并能同时融入时、空信息,更适用于受人类推理、判断、决策等思维过程启发的类脑计算.因此,介绍了脉冲神经网络的特点、脉冲神经元模型、脉冲编码,以及脉冲神经网络在模式识别等领域中的应用,并对基于脉冲神经网络的类脑计算方法和神经形态芯片的研究现状及未来的发展进行了讨论.基于脉冲神经网络的类脑计算会对未来的生活、经济发展产生深远影响.

       

      Abstract: Nowadays, brain-inspired computing is one of the most important international research frontiers of science and technology. The paper is focused on the topic about brain-inspired computing. The features, research contents and situation of brain-inspired computing were discussed. Brain-inspired computing is not only the simple simulation of neurons in human being's brains, but also the simulation for the rules of people's information processing, complex working mode, and the essence nature of human being's thought, study, deduction and decisions making. As brain-inspired computing is enlightened by the process of human's reasoning, judging, and decisions making, pulsed neural networks (pulsed NNs) incorporated with temporal and spatial information simultaneously, which are bio-plausible and different from the traditional artificial neural networks, are suitable for brain-inspired computing. Pulsed NNs' characters, neuron models, and pulsed encoding and the applications in pattern recognition and other disciplines were presented in this paper. The situation and future direction of the research of brain-inspired computing methods and neuromorphic chips based on pulsed NNs were discussed. Brain-inspired computing based on pulsed NNs will significantly affect our life and economy in the future.

       

    /

    返回文章
    返回