矿山信息物理融合系统多节点智联策略

Intelligent multi-node communication strategy of mine cyber-physical system

  • 摘要: 针对当前矿山信息物理融合系统(CPS)的通信节点无法与基于不同无线通信协议的感知节点实现智能连接的问题,在通信节点上集成多种通信模块构成多模态通信节点,提出了一种基于渐进式神经网络的矿山CPS多节点智联策略。采用渐进式神经网络控制多模态通信节点准确切换工作模态,实现异构无线通信网络自主建立;利用异步优势动作评价算法对渐进式神经网络进行深度训练,提高渐进式神经网络的收敛速度和训练精度。实验结果表明,该策略实现了多模态通信节点与多类感知节点之间的准确、可靠通信。

     

    Abstract: Aiming at problem that communication nodes and perception nodes based on different wireless communication protocols could not achieve intelligent connection in current mine cyber-physical system (CPS), a multi-mode communication node was constructed by integrating multiple communication modules on the communication node, and an intelligent multi-node communication strategy of mine CPS based on progressive neural network was proposed. The progressive neural network is used to control the multi-mode communication node to switch working mode accurately and realize independent establishment of heterogeneous wireless communication network. The asynchronous advantage actor-critic algorithm is used to perform deep training on the progressive neural network to improve convergence speed and training accuracy of the progressive neural network. The experimental results show that the strategy can realize accurate and reliable communication between multi-mode communication nodes and multi-class perception nodes.

     

/

返回文章
返回