船舶通信网络异常数据自动检测和剔除方法  

Research on automatic detection and elimination methods for abnormal data in ship communication networks

作　　者：侯立 杨成佳[2] HOU Li;YANG Cheng-jia(College of Computer Science and Technology,Jilin Normal University,Changchun 130022,China;School of Computing,Jilin Jianzhu University,Changchun 130022,China)

机构地区：[1]吉林师范大学计算机科学与技术学院,吉林长春130022 [2]吉林建筑大学计算机学院,吉林长春130022

出　　处：《舰船科学技术》2023年第19期173-176,共4页Ship Science and Technology

基　　金：吉林省教育科学规划课题(GH21011)。

摘　　要：为提高船舶通信网络异常数据自动检测精度,并全面剔除船舶通信网络异常数据,研究新的船舶通信网络异常数据自动检测和剔除方法。利用基于改进支持向量机的网络异常数据自动检测方法,由改进粒子群优化算法,寻优设置支持向量机的惩罚项、核函数的预定义参数,训练性能合格的支持向量机后,以船舶通信网络数据分类的方式,自动检测船舶通信网络异常数据;将异常数据使用基于自适应级联陷波器的异常数据剔除方法,通过自适应级联陷波器,以异常数据滤波的方式,剔除船舶通信网络异常数据。研究结果显示,使用所提方法,船舶通信网络异常数据自动检测结果符合实际数目,可有效去除船舶通信网络异常数据。To improve the accuracy of automatic detection of abnormal data in ship communication networks and comprehensively eliminate abnormal data in ship communication networks,a new method for automatic detection and elimination of abnormal data in ship communication networks is studied.This method utilizes an improved support vector machine based network anomaly data automatic detection method.The improved particle swarm optimization algorithm optimizes and sets the penalty terms and predefined parameters of the kernel function of the support vector machine.After training a qualified support vector machine,the ship communication network anomaly data is automatically detected through classification of ship communication network data;Using an adaptive cascaded notch filter based anomaly data removal method,the abnormal data in the ship communication network is filtered through the adaptive cascaded notch filter.The research results show that under the use of the proposed method,the automatic detection results of abnormal data in the ship communication network match the actual number,and can effectively remove abnormal data in the ship communication network.

关 键 词：船舶通信网络 异常数据 自动检测 剔除方法 

分 类 号：TP393[自动化与计算机技术—计算机应用技术]