基于人工智能技术的鲲鹏云服务系统可靠性评估方法  

Reliability Evaluation Method of Kunpeng Cloud Service System Based on Artificial Intelligence Technology

作　　者：曾赟[1] ZENG Yun(College of Information Engineering,Yellow River Conservancy Technical Institute,Kaifeng 475000,China)

机构地区：[1]黄河水利职业技术学院,信息工程学院,河南开封475000

出　　处：《微型电脑应用》2023年第6期58-61,共4页Microcomputer Applications

基　　金：黄河水利职业技术学院教育教学改革研究项目(2020XJJGLXZB008)。

摘　　要：针对目前鲲鹏云服务系统可靠性评估精确度不高、及时性差等问题,提出基于人工智能技术的鲲鹏云服务系统可靠性评估方法。该方法明确系统可靠性相关概念,建立鲲鹏云服务系统的不同阶段运行故障模型;运用Delta学习方法调整权重来优化神经网络性能,将训练数据归一化为零均值;在反向误差传播后持续进行前向传播直至误差收敛,误差小于收敛值则结束训练,获得可靠性评估输出值。实例测试结果表明,所提方法可有效处理鲲鹏云服务系统庞大的数据集合,可靠性评估误差低至0.1%,精度较高,收敛时间不超过80 s,计算速度快,鲁棒性强。Aiming at the problems of low accuracy and poor timeliness in reliability evaluation of Kunpeng cloud service system,a reliability evaluation method based on artificial intelligence technology is proposed.The relevant concepts of system reliability are clarified,and the weight is adjusted by delta learning method to optimize the performance of neural network.The training data are normalized to zero mean value,and the forward propagation is continued until the error converges after the backward error propagation.When the error is less than the convergence value,the training is ended,and the output value of reliability evaluation is obtained.The test results show that the proposed method can effectively deal with the huge data set of Kunpeng cloud service system,the reliability evaluation error is about 0.1%,and the accuracy is high,the convergence time is less than 80 s,the calculation speed is fast,and the robustness is strong.

关 键 词：人工智能技术 鲲鹏产业 云服务系统 可靠性评估 人工神经网络 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]