飞轮系统的稳定运行对于航天器在轨安全影响重大，因而对飞轮系统进行健康状态评估至关重要。在进行飞轮系统健康状态评估建模时，不仅要求模型能够处理各种不确定性以保障评估结果的准确性，同时要求其具有透明合理的评估过程与可解释、可追溯的评估结果。因此，本文在深入研究置信规则库(Belief Rule Base, BRB)建模方法的基础上，构建了一种新的基于可解释性建模的置信规则库（Explicable Belief Rule Base， BRB-e）飞轮系统健康状态评估模型。首先，结合飞轮系统特征对模型的可解释性建模准则进行定义；在此基础上，设计了BRB-e评估模型的推理过程；然后，基于鲸鱼优化算法（Whale Optimization Algorithm，WOA），提出了一种具有可解释性约束的BRB-e模型参数优化方法；最后，通过对某飞轮系统中轴承组件的评估案例研究，验证了模型在飞轮系统健康状态评估中的有效性。对比研究表明，BRB-e模型在评估结果准确性和评估过程可解释性方面具有一定的优势。
Abstract: The stable operation of the flywheel system has a great impact on the on orbit safety of spacecraft, so it is very important to assess the health status of the flywheel system. When modeling the flywheel system health sta-tus assessment, it is required that the model not only deal with various uncertainties to ensure the accuracy of the assessment results, but also have a transparent and reasonable assessment process and explicable and traceable assessment results. Therefore, based on the in-depth study of the modeling method of belief rule base (BRB), this paper constructs a new explicable belief rule base (BRB-e) flywheel system health assessment model based on expli-cable modeling. Firstly, the explicable modeling criteria are defined according to the characteristics of flywheel system; On this basis, the reasoning process of BRB-e assessment model is designed; Then, based on the whale optimization algorithm (WOA), a parameter optimization method of BRB-e model with explicable constraints is proposed; Finally, the effectiveness of the model in the flywheel system health status assessment is verified by a case study of bearing components in a flywheel system. The comparative study shows that BRB-e model has certain advantages in the ac-curacy of assessment results and the explainability of assessment process.
[1]GHANAATIAN M and LOTFIFARD S.Control of flywheel energy storage systems in the presence of uncertain-ties[J].IEEE Transactions on Sustainable Energy, 2018, 10(1):36-45
[2]SUN XD, SU BK, WANG SH, et al.Performance analysis of suspension force and torque in an IBPMSM with V-shaped PMs for flywheel batteries[J].IEEE Transactions on Magnetics, 2018, 54(11):1-4
[3]ZHOU ZJ, CAO Y, HU GY, et al.New health-state assess-ment model based on belief rule base with interpretabil-ity[J].Science China Information Sciences, 2021, 64(7):1-15
[4]ELSHEIKH AH, SHARSHIR SW, ABD ELAZIZ M, et al.Modeling of solar energy systems using artificial neural net-work: A comprehensive review[J].Solar Energy, 2019, 180:622-639
[5]SPEISER JL, MILLER ME, TOOZE J, et al.A comparison of random forest variable selection methods for classification prediction modeling[J].Expert systems with applications, 2019, 134:93-101
[6]MU YS, LIU XD and WANG LD.A Pearson’s correlation coefficient based decision tree and its parallel implementa-tion[J].Information Sciences, 2018, 435:40-58
[7]YANG JB, LIU J, WANG J, et al.Belief rule-base inference methodology using the evidential reasoning approach-RIMER[J].IEEE Transactions on systems, Man, and Cybernet-ics-part A: Systems and Humans, 2006, 36(2):266-285
[8]YANG JB and XU DL.Evidential reasoning rule for evidence combination[J].Artificial Intelligence, 2013, 205:1-29
[9]CAO Y, ZHOU ZJ, HU CH, et al.On the interpretability of Belief Rule-Based expert systems[J].IEEE Transactions on Fuzzy Systems, 2021, 29(11):3489-3503
[10]周志杰, 曹友, 胡昌华等.基于规则的建模方法的可解释性及其发展[J].自动化学报, 2021, 47(6):1201-1216
[11]GAO Y, LIU XY and XIANG JW.FEM simulation-based generative adversarial networks to detect bearing faults[J].IEEE Transactions on Industrial Informatics, 2020, 16(7):4961-4971
[12]DENG XY and JIANG W.Dependence assessment in human reliability analysis using an evidential network approach ex-tended by belief rules and uncertainty measures[J].Annals of Nuclear Energy, 2018, 117:183-193
[13]CHEN X, CHENG L, LIU C, et al.A woa-based optimization approach for task scheduling in cloud computing systems[J].IEEE Systems journal, 2020, 14(3):3117-3128
[14]YANG JB.Rule and utility based evidential reasoning ap-proach for multi-attribute decision analysis under uncertain-ties[J].European Journal of Operational Research, 2001, 131(1):31-61
[15]ZHOU ZJ, CAO Y, HU GY, et al.. 2021; 64(7): 15.[J].Science China Information Sciences, 2021, 64(7):15-
[16]LI GL, ZHOU ZJ, HU CH, et al.optimal safety assess-ment model for complex systems considering correlation and redundanc[J].International Journal of Approximate Reasoning, 2019, 104:38-56
[17]LI ZM, BEJARBANEH BY, ASTERIS PG, et al.A hybrid GEP and WOA approach to estimate the optimal penetration rate of TBM in granitic rock mass[J].Soft Computing, 2021, 25(7):11877-11895
[18]KONG DD, CHEN YJ, LI N, et al.Tool wear estimation in end milling of titanium alloy using NPE and a novel WOA-SVM model[J].IEEE Transactions on Instrumentation and Measurement, 2019, 69(7):5219-5232
[19]ZHAO HM, LIU HD, XU JJ, et al.Performance prediction using high-order differential mathematical morphology gra-dient spectrum entropy and extreme learning machine[J].IEEE Transactions on Instrumentation and Measurement, 2019, 69(7):4165-4172
[20]SONG HH, WEN H, HU L, et al.Secure cooperative trans-mission with imperfect channel state information based on BPNN[J].IEEE Transactions on Vehicular Technology, 2018, 67(11):10482-10491
CJA