摘要
针对传统电压频控软件缺陷检测技术未考虑软件缺陷分类,存在检测精度低的问题,提出一种电压频控中抗强干扰软件关联缺陷检测技术;对软件关联缺陷检测原理进行分析,采用判别函数对待测软件样本进行识别,引入统计模式识别算法处理软件原始数据,依据关联缺陷概率分配,确定关联缺陷类别,计算缺陷特征值,利用贝叶斯分类器对关联缺陷进行划分,完成抗强干扰软件关联缺陷的分类,从而实现关联缺陷的高精度检测;实验结果表明,该检测技术对软件缺陷进行准确分类,在保证强抗干扰性的前提下,有效提高了检测精度。
The traditional voltage and frequency control software defect detection technology does not consider the classification of software defects,and has the problem of low detection accuracy.The software defect correlation detection principle were analyzed using discriminant function test software to identify sample treatment,introduced statistical pattern recognition algorithm of original data processing software,based on defect correlation probability distribution,determine the associated defects of calculation of defect feature values to classify the defect correlation using Bayesian classifier,complete classification of anti interference software defect correlation in order to achieve high precision detection,defect correlation.The experimental results show that the detection technology can accurately classify the software defects and effectively improve the detection accuracy on the premise of ensuring strong anti-interference.
作者
刘欣
张楠
孙辰军
Liu Xin;Zhang Nan;Sun Chenjun(State grid hebei electirc power company,Shijiazhuang 050021,China;Nari group corporation nanjing,Nanjing 210000,China)
出处
《计算机测量与控制》
2018年第4期11-14,共4页
Computer Measurement &Control
基金
国家电网公司总部科技项目(XX71-15-036)
关键词
电压频控
抗强干扰
软件缺陷
检测
voltage frequency control
anti strong interference
software defect
detection
