计算机科学 ›› 2023, Vol. 50 ›› Issue (5): 31-37.doi: 10.11896/jsjkx.220900283
李红辉1,2, 陈博1, 鲁姝艺1, 张骏温1
LI Honghui1,2, CHEN Bo1, LU Shuyi1, ZHANG Junwen1
摘要： 软件可靠性预测以软件可靠性预测模型为基础,对软件的可靠性以及与其直接相关的度量进行分析、评价和预测,利用软件运行中所收集的失效数据对未来的软件可靠性进行预测,成为了评估软件失效行为和保障软件可靠程度的重要手段。BP神经网络结构简单、参数少、易实现,在软件可靠性预测领域已经得到了广泛应用。然而基于传统BP神经网络搭建的软件可靠性预测模型的预测精度无法达到预期目标,因此提出了基于BASFPA-BP的软件可靠性预测模型。该模型利用软件失效数据,在BP神经网络训练过程中利用BASFPA算法优化网络权值、阈值,从而提高模型的预测精度。选用3组公开的软件失效数据,将实际值与预测值的均方误差作为预测结果的衡量标准,同时将BASFPA-BP与FPA-BP,BP,Elman这3种模型进行对比研究。实验结果表明,基于BASFPA-BP的软件可靠性预测模型在同类型模型中实现了较高的预测精度。
中图分类号:
| [1]NIKORA A,FIONDELLA L,WANDJI T.SFRAT-An Exten-dable Software Reliability Assessment Tool[C]//2018 IEEE International Symposium on Software Reliability Engineering Workshops(ISSREW).IEEE,2018:65-70. [2]BAL P R,MOHAPATRA D P.Software reliability prediction based on radial basis function neuralnetwork[M].Singapore:Springer,2017:101-110. [3]JELINSKI Z,MORANDA P.Software reliability research[M]//Statistical Computer Performance Evaluation.Academic Press,1972:465-484. [4]MUSA J D,OKUMOTO K.A logarithmic Poisson executiontime model for software reliability measurement[C]//Procee-dings of the 7th International Conference on Software Enginee-ring.1984:230-238. [5]OKUMOTO K,GOEL A L.Availability and other performancemeasures of software systems under imperfect maintenance[C]//The IEEE Computer Society's Second International Computer Software and Applications Conference(COMPSAĆ78).IEEE,1978:66-70. [6]LITTLEWOOD B,VERRALL J L.A Bayesian reliability growthmodel for computer software[J].Journal of the Royal Statistical Society:Series C(Applied Statistics),1973,22(3):332-346. [7]DUANE J T.Learning Curve Approach to Reliability Monitoring[J].IEEE Transactions on Aerospace,1964,2(2):563-566. [8]YAMADA S,OHBA M,OSAKI S.S-Shaped Reliability Growth Modeling for Software Error Detection[J].IEEE Transactions on Reliability,1983,32(5):475-484. [9]LIU L,HA J H,WANG A B.Software reliability growth model based on FABP[J].Computer simulation,2015,32(6):440-446. [10]WANG J,ZHANG C.Software reliability prediction using adeep learning model based on the RNN encoder-decoder[J].Reliability Engineering & System Safety,2018,170:73-82. [11]KUZNETSOVV,MARIET Z.Foundations of Sequence-to-Se-quence Modeling for Time Series[C]//Machine Learning Research.2019:408-417. [12]YAN A,GU J,MU Y,et al.Research on photovoltaic ultrashort-term power prediction algorithm based on attention and LSTM[C]//IOP Conference Series:Earth and Environmental Science.IOP Publishing,2021. [13]DUBEY H M,PANDIT M,PANIGRAHI B K.A Biologically Inspired Modified Flower Pollination Algorithm for Solving Economic Dispatch Problems in Modern Power Systems[J].Cognitive Computation,2015,7(5):594-608. [14]HE X,YANG X S,KARAMANOGLU M,et al.Global Conver-gence Analysis of the Flower PollinationAlgorithm:A Discrete-Time Markov Chain Approach[J].Procedia Computer Science,2017,108:1354-1363. [15]BIAN H J,HE X S,YANG X S.Adaptive pollination optimization algorithm based on Firefly algorithm[J].Computer Engineering and Applications,2016,52(21):162-167,217. [16]XIAO H H,WANG C X,DUAN Y M,et al.Flower pollination optimization algorithm based on Gaussian mutation and Powell method[J].Computer Science and Exploration,2017,11(3):478-490. [17]JIANG X,LI S.BAS:Beetle Antennae Search Algorithm forOptimization Problems[J].arXiv:1701.10724,2017. [18]SHAO L S,HAN R D.Flower Pollination algorithm based on Longicorn beard search[J].Computer Engineering and Applications,2018,54(18):188-194. [19]LU S Y.Research on Software reliability prediction model based on LSTM[D].Beijing:Beijing Jiaotong University,2021. |
|
||
首页