摘要
智能汽车、无人机或工业智能化生产线,在边缘计算提供的共享式传感网和共享式算力资源帮助下,能够有效降低成本,扩大感知范围。传统的通信网络虽然实现了全球统一的广域连接,但在动态更新局部域通信过程中,远距离的传感器对于受控对象所处空间控制的意义并不大。为了提高区域通信的实时响应,通信网络需要随着汽车、无人机或生产线物流的空间位置变化进行动态局部组网。基于此,本文介绍了一种新型的边缘计算分布式通信网络,提出了动态更新域标识架构,该架构内的各终端节点根据空间所处位置自动地构造局域通信簇,并基于该簇进行动态轮询访问。该方法能够实现近距共享传感器和人工智能算力,降低了单机智能化成本,对于人工智能项目落地和技术推广具有一定的促进意义。
With the help of shared sensor networks and shared computing resources provided by edge computing,intelligent driverless cars,unmanned aerial vehicles or industrial intelligent production lines can effectively reduce costs and expand the range of sensors perception.Although the traditional internet has realized the global unified wide area connection,in the process of dynamic network address for updating local area communication,the remote sensors are useless for controlling objects compared to local sensors.In order to improve the real-time response of regional communication about controllers and controlled object devices,the network needs dynamic local connecting with the spatial location changes of cars,unmanned aerial vehicles or production line logistics.Based on this,a new type of edge computing distributed communication network and a dynamically updated domain identity architecture have been proposed in this paper.Each endpoint node can automatically construct a local cluster according to its location in space for signal transfer,and dynamically polls and accesses the cluster based on the cluster.It can realize close sharing of sensors and Al computing power and reduce the cost of building a comprehensive intelligent control system compare to traditional singlemachine Al systems.It has certain promotion significance for the implementation of Al comprehensive projects.
出处
《自动化博览》
2023年第2期62-66,共5页
Automation Panorama1
基金
云南省第七批高校重点实验室和工程研究中心建设项目,云南省科技厅重点研发计划(202103AC100002)。
