基于马尔科夫链的软件系统接口预警技术研究

The Research on Software System Interface Early Warning Technology Based on Markov Chain

文章提出并实现了一种软件接口评估及故障预警技术,旨在获取软件接口连接错误的故障层级标准,并通过分析软件接口原始日志中的报错信息来确定接口错误的故障层级。研究采用了时间窗的概念,对各故障层级在时间窗内的发生频次进行统计,并将其作为输入数据,输入到基于马尔可夫链的软件接口评估及故障预警模型中,该模型能够输出软件接口故障的预测及评价,从而提高故障预警和接口评估的整体标准及可靠性。文章提出的技术优点在于其适用性广泛,可以适应多种不同的应用场景,并且其原理亦可应用于多场合的时间序列预测。

This paper proposes and implements a software interface evaluation and fault prediction technology aimed at obtaining a fault hierarchy standard for software interface connection crrors.It determines the fault hierarchy of interface crrors by analyzing error information in the original log of software interfaces.The study cmploys the concept of a time window to statistically calculate the occurrence frequency of each fault level within the time window.These frequcncics are uscd as input data for a software interface evaluation and fault prediction model based on the Markov chain.The model is capable of outputting predictions and evaluations of software interface failures,thereby enhancing the overall standards and reliability of fault prediction and interface evaluation.The advantage of the technology proposed in this paper is its wide applicability,which can adapt to various different application scenarios,and its principles can also be applied to time series prediction in multiple situations.