摘要
对于软件系统的可靠性模型而言,其主要功能就是要求能准确地预测出后续的软件失效的发生时间,研究了软件失效时序数据的灰色拟合问题,并提出了相应的多步预测改进算法。首先,基于某一拟合精度的条件下,应用灰色理论的GM(1,1)模型为软件失效时序数据建立相应的分段拟合模型;然后,以正交多项式函数估算出末段子序列的拟合残差,并用该估计值对末段子序列的灰色预测模型进行修正和补偿;最后,基于补偿后的组合预测模型对软件失效的发生时间进行多步预测。实验验证了算法的正确性和有效性。
For the reliability model of the software system,its main function is to accurately predict the occurrence of a subsequent software failure time. The paper studies grey fitting problem of the software failure time series data,and proposes the corresponding multi-step prediction algorithm. First of all,based on the condition of a fitting accuracy,apply grey theory GM（ 1,1） model for the software failure time series data to establish the corresponding piecewise fitting model; then,use the orthogonal polynomial function to estimate fitting residual error at the end of the sequence,and utilize the estimate result for correcting and compensating grey prediction model at the end of the sequence; in the end,apply the combined forecasting model based on compensation to realize multi-step projections for the occurrence of software failure time. The experiment verifies the correctness and effectiveness of the algorithm.
出处
《智能计算机与应用》
2017年第6期9-13,16,共6页
Intelligent Computer and Applications
基金
佛山职业技术学院校级重点科研项目(2015KY006)
