首页|基于脉冲神经网络的机器人智能控制研究进展

基于脉冲神经网络的机器人智能控制研究进展

扫码查看
像人一样在复杂多变的非结构化环境中灵巧操作是机器人研究追求的目标之一.受生物脑工作方式启发的脉冲神经网络(SNN)是类脑智能领域的主要工作范式,具有良好的生物合理性,在机器人智能控制领域日益受到关注.本文对基于SNN的机器人类脑智能控制相关研究展开综述,期望能为机器人和类脑智能领域的研究带来启发.首先,介绍SNN的发展历程、神经元模型、编码方式、突触可塑性和网络结构等相关知识.其次,借鉴人类的运动反馈控制机制,给出一种基于SNN的机器人类脑智能控制框架.再次,从运动控制、柔顺控制、协同控制三个方面介绍机器人类脑智能控制策略的研究进展.最后,对基于SNN的机器人类脑智能控制技术进行了总结与展望.
Research Advances on Robot Intelligent Control Based on Spiking Neural Network
One of the goals pursued by robotics research is to operate dexterously like humans in complex and changeable unstructured environments.The spiking neural network(SNN)inspired by the working mode of the biological brain is the main working paradigm in the field of brain-inspired in-telligence.It has good biological rationality and has attracted increasing attention in the field of ro-bot intelligent control.We provide a review of research related to SNN-based robot brain-inspired intelligent control,hoping to bring inspiration to research in the field of robots and brain-inspired intelligence.Firstly,we introduce SNN-related knowledge such as the development history of SNN,neuron models,encoding methods,synaptic plasticity and network structure.Secondly,by drawing on the human motion feedback control mechanism,we give a framework for robot brain-in-spired intelligent control based on SNN.Thirdly,we introduce the research progress of robot brain-inspired intelligent control strategies from three aspects:motion control,compliance control,and collaborative control.Finally,the SNN-based robot brain-inspired intelligent control technology is summarized and prospected.

spiking neural networkmotion controlcompliance controlcollaborative control

敖天勇、曹贤泽、付乐、周毅

展开 >

河南大学人工智能学院,河南郑州 450046

河南省车联网协同技术国际联合实验室,河南郑州 450046

脉冲神经网络 运动控制 柔顺控制 协同控制

国家自然科学基金项目国家自然科学基金项目河南省科技攻关项目河南省科技攻关项目

6217608862303160232102211011242102220111

2024

信息与控制
中国自动化学会 中国科学院沈阳自动化研究所

信息与控制

CSTPCD北大核心
影响因子:0.576
ISSN:1002-0411
年,卷(期):2024.53(4)