Hard disk detection method and device, storage medium and equipment

Abstract

The embodiment of the invention discloses a hard disk detection method and device, a storage medium and a terminal, which is applied to the technical field of hardware security. The method comprises the steps of obtaining hard disk data; obtaining binary classification output of the hard disk through an optimal support vector machine model according to the hard disk data, and judging whether the hard disk is a first type hard disk or not; if yes, obtaining a risk coefficient of the first type of hard disk, wherein the risk coefficient is obtained according to hard disk data; and inputting thehard disk data of the first type of hard disk and the risk coefficient into a risk estimation model to obtain a damage risk value of the hard disk. And the damage risk of the hard disk can be judged according to the specific damage risk value, thereby being more accurate and visual.

Description

technical field [0001] The invention relates to the field of computer technology, in particular to a hard disk detection method, device, storage medium and equipment. Background technique [0002] A hard disk is a computer storage device that integrates precision machinery, microelectronic circuits, and electromagnetic conversion. It stores computer system resources and important information and data. These factors make the hard disk one of the most important hardware devices in a PC. [0003] For the detection of hard disks, an existing detection method is the threshold judgment method generally used in the industry, that is, online monitoring of changes in the S.M.A.R.T threshold, and an alarm will be issued if the threshold is exceeded. Another existing hard disk detection method uses cloud computing platform hard disk failure prediction, and according to the hard disk repair records within the prediction time window, the hard disk S.M.A.R.T log data is marked as normal h...

AI Technical Summary

Problems solved by technology

[0004] Since the threshold in the threshold determination method is set by the manufacturer, in order to reduce the number of hard disks returned to the factory for inspection and repair due to false alarms, the manufacturer will set the threshold quite strictly. Although the false alarm rate can be as low as 0.1%, it will result in only Very few hard drive failures detected

In addition, when using S.M.A.R.T data, the failure-related features are weakened due to improper data selection, and different types of hard drives have different ranges of the same characteristics, so missing values ​​cannot be simply filled, which will lead to serious deviations in the prediction model

The voting mode is easy to lead to misjudgment, that is, using the models established on multiple "balanced" data sets to vote for whether to judge whether the fault is judged, it is easily affected by the distribution of the data set, and it is obtained by voting through K support vector machine classifier models Binary judgment of whether each disk is about to fail, but does not evaluate the risk of the disk about to fail, and cannot provide the owner of the hard disk with an inference of the failure probability within a period of time

Images