基于卷积神经网络的智能机器人信息安全访问控制方法

Control method of intelligent robot based on convolutional neural network

针对智能机器人接收的多样模态访问信息,文章提出了基于卷积神经网络的访问控制方法。通过量化处理图像、文本、语音数据,引入卷积神经网络分析“超平面”中的映射分布,确定访问控制函数输出。测试显示,该方法下异常访问对系统安全态势影响小,参数稳定在90.0以上,显著优于对照组,展现了可靠性方面的优势。

This article proposes an access control method based on convolutional neural networks for the diverse modal access information received by intelligent robots.By quantifying image,text,and speech data,a convolutional neural network is introduced to analyze the mapping distribution in the hyperplane and determine the output of the access control function.Tests have shown that abnormal access under this method has little impact on the system security situation,with parameters stable above 90.0,significantly better than the control group,demonstrating advantages in reliability.