基于智能学习模型的民航发动机健康状态预测研究
|
摘要
民航发动机健康状态预测是制定合理的发动机调度和维修规划的基础，也是保障运营安全、提高经济性的重要支撑技术。作为无需先验假设的数据驱动模型，以神经网络为代表的智能学习模型可直接利用发动机健康状态数据建立预测模型，解决了发动机健康状态衰退解析模型难以建立的问题。但是，发动机健康状态衰退过程的复杂非线性时变特性给智能学习模型的模型选择和优化带来了较大困难。
本文针对具有复杂非线性时变特性的民航发动机健康状态预测的建模和求解问题，对自适应数据降噪、单一全局预测模型和集成局域预测模型及模型预测结果不确定性评估等健康状态预测技术进行了系统研究。
提出一种基于经验模态分解增强的奇异值分解健康状态信号降噪方法，采用经验模态分解从原始健康状态信号提取趋势分量，对信号剩余部分进行奇异值分解降噪，消除趋势分量对奇异值差分谱的干扰，实现信号重构奇异值的自适应选择。发动机健康状态信号降噪结果证明了该方法的有效性。
针对目前连续函数输入型过程神经网络不能直接利用离散样本进行训练的不足，提出一种离散输入过程神经网络模型。离散输入过程神经网络以离散向量作为输入，以卷积和算子实现时间聚合运算，可以避免连续函数输入过程神经网络在函数拟合和基展开过程中存在的参数难以选择及信息丢失的问题。民航发动机健康状态预测实例表明，离散输入过程神经网络在预测精度不低于连续函数输入过程神经网络情况下，具有更好的操作性。
针对单一全局预测模型存在模型复杂难以优化的问题，提出了基于改进AdaBoost.RT的静态权值和动态权值组合集成模型。对AdaBoost.RT算法的误差函数进行了改进，并采用自适应阈值调整方法在训练过程中自动调整分类阈值。分别以离散输入过程神经网络、传统人工神经网络和极端学习机为弱学习机构建静态权值和动态权值组合集成预测模型。其中，动态权值集成模型在训练过程中采用训练样本的近邻样本对弱学习机进行性能评估，根据学习机在测试样本近邻样本上的性能来动态赋予学习机组合权值，从而可以充分挖掘弱学习机的局部性能，进一步提高了预测效果。对民航发动机的健康状态预测结果表明，由具有简单结构弱学习机构建的集成预测模型的预测效果好于单一全局模型，且集成模型可以降低对弱学习机的性能要求。
针对模型点预测结果存在的不确定性，采用Bootstrap预测区间估计方法对所提出的上述预测模型的预测区间进行估计，实现对模型预测结果可靠性和精确性的量化评估。
基于上述研究成果，根据中国国际航空股份有限公司的实际需求研发了“民航发动机拆发日期预测系统”，成为“航空发动机健康管理与维修决策支持系统”的重要组成模块。该模块从系统获得发动机的健康状态信息实现发动机拆发日期预测，将预测结果提供给系统的“维修计划制定”模块作为维修计划制定的依据，实现了从发动机状态监控到维修决策的无缝集成。
The health condition prediction of civil aircraft engine is the basis for makingreasonable scheduling and maintenance plans, and is the support technology forensuring operation safety and improving economy. As data-driven models withoutprior hypotheses, intelligent learning models represented by neural networks can betrained using the health condition data directly, which overcomes the difficulties ofestablishing exact mechanism models for engine health condition prediction.However, the complex nonlinear time-varying characteristics of the aircraft enginehealth degradation process bring about difficulties to the actual use of intelligentlearning models.
In view of such problems and the requirements of civil aircraft engine healthcondition prediction, the adaptive noise reducing method, the single global modelingmethod, the ensemble local modeling method and the uncertainty assessing methodfor prediction results are studied in this paper.
An engine health signal noise reducing method based on singular valuedecomposition (SVD) enhanced by empirical mode decomposition (EMD) isproposed. The trend component is abstracted from the original health signal usingEMD, and then the residual is de-noised using SVD. Since the disturbance of thetrend component is eliminated, the singular values for signal reconstruction can beselected adaptively according to the singlular value spectrum. The given method isvalidated through the noise reducing of the actual engine health signals.
In view of the problem that the process neural network (PNN) with continuousfunction inputs can not be trained using discrete samples directly, a discrete inputprocess neural network model (DPNN) is presented. DPNN takes discrete vectors asinputs and utlize convolution sum to realize the time aggregate operation. Thus,DPNN can avoid problems such as parameter selection and information loss duringprocedures of function fitting and expanding required by PNN with continuousfunction inputs. The engine health condition prediction results show that, DPNNperforms comparable to PNN with continuous function inputs while with betteroperability.
To overcome the difficulties of model optimization required by single globalmodels for engine health condition prediction, two ensemble prediction models with static and dynamic combining weights respectively based on the improvedAdaBoost.RT are proposed. The error function of AdaBoost.RT is improved, and anadaptive adjustment strategy is adopted to adjust the threshold during the trainingprocess. Then, DPNN, ANN and the extreme learning machine are utilized as weaklearners to construct ensemble models for engine health condition prediction. Forensemble models with dynamic combining weights, the weak learners are evaluatedusing the neighboring samples during the training process, and the dynamiccombining weights of the weak learners are generated according to their performanceon the neighboring samples of the testing sample in the training sample set. Since thelocal performance of the weak learners can be well mined, the ensemble models withdynamic combining weights are superior to their counterparts with static combiningweights. The prediction results of engine health conditions show that the ensemblemodels always perform better than single global models. And the performancerequirements for predictors can be reduced while being utilized as weak leaners.
According to the fact that the point prediction results are accompanied withuncertainty, the Bootstrap based prediction interval estimation method is employed toestimate the prediction intervals of the presented models, thus to achieve quantitativeassessment of the reliability and accuracy of the prediction results.
Based on such reseach, the “system for engine removal date prediction” isdeveloped according to the actual requirements of Air China. As a key subsystem ofthe “aircraft engine health management and maintenance decision support system”,the developed system acquires health condition information from its parent system toexecute engine removal date prediction, and the results are provided to themaintenance planning module for maintenance plan making, which implementsseamless integration from engine condition monitoring to maintenance decisionmaking.
本文针对具有复杂非线性时变特性的民航发动机健康状态预测的建模和求解问题，对自适应数据降噪、单一全局预测模型和集成局域预测模型及模型预测结果不确定性评估等健康状态预测技术进行了系统研究。
提出一种基于经验模态分解增强的奇异值分解健康状态信号降噪方法，采用经验模态分解从原始健康状态信号提取趋势分量，对信号剩余部分进行奇异值分解降噪，消除趋势分量对奇异值差分谱的干扰，实现信号重构奇异值的自适应选择。发动机健康状态信号降噪结果证明了该方法的有效性。
针对目前连续函数输入型过程神经网络不能直接利用离散样本进行训练的不足，提出一种离散输入过程神经网络模型。离散输入过程神经网络以离散向量作为输入，以卷积和算子实现时间聚合运算，可以避免连续函数输入过程神经网络在函数拟合和基展开过程中存在的参数难以选择及信息丢失的问题。民航发动机健康状态预测实例表明，离散输入过程神经网络在预测精度不低于连续函数输入过程神经网络情况下，具有更好的操作性。
针对单一全局预测模型存在模型复杂难以优化的问题，提出了基于改进AdaBoost.RT的静态权值和动态权值组合集成模型。对AdaBoost.RT算法的误差函数进行了改进，并采用自适应阈值调整方法在训练过程中自动调整分类阈值。分别以离散输入过程神经网络、传统人工神经网络和极端学习机为弱学习机构建静态权值和动态权值组合集成预测模型。其中，动态权值集成模型在训练过程中采用训练样本的近邻样本对弱学习机进行性能评估，根据学习机在测试样本近邻样本上的性能来动态赋予学习机组合权值，从而可以充分挖掘弱学习机的局部性能，进一步提高了预测效果。对民航发动机的健康状态预测结果表明，由具有简单结构弱学习机构建的集成预测模型的预测效果好于单一全局模型，且集成模型可以降低对弱学习机的性能要求。
针对模型点预测结果存在的不确定性，采用Bootstrap预测区间估计方法对所提出的上述预测模型的预测区间进行估计，实现对模型预测结果可靠性和精确性的量化评估。
基于上述研究成果，根据中国国际航空股份有限公司的实际需求研发了“民航发动机拆发日期预测系统”，成为“航空发动机健康管理与维修决策支持系统”的重要组成模块。该模块从系统获得发动机的健康状态信息实现发动机拆发日期预测，将预测结果提供给系统的“维修计划制定”模块作为维修计划制定的依据，实现了从发动机状态监控到维修决策的无缝集成。
The health condition prediction of civil aircraft engine is the basis for makingreasonable scheduling and maintenance plans, and is the support technology forensuring operation safety and improving economy. As data-driven models withoutprior hypotheses, intelligent learning models represented by neural networks can betrained using the health condition data directly, which overcomes the difficulties ofestablishing exact mechanism models for engine health condition prediction.However, the complex nonlinear time-varying characteristics of the aircraft enginehealth degradation process bring about difficulties to the actual use of intelligentlearning models.
In view of such problems and the requirements of civil aircraft engine healthcondition prediction, the adaptive noise reducing method, the single global modelingmethod, the ensemble local modeling method and the uncertainty assessing methodfor prediction results are studied in this paper.
An engine health signal noise reducing method based on singular valuedecomposition (SVD) enhanced by empirical mode decomposition (EMD) isproposed. The trend component is abstracted from the original health signal usingEMD, and then the residual is de-noised using SVD. Since the disturbance of thetrend component is eliminated, the singular values for signal reconstruction can beselected adaptively according to the singlular value spectrum. The given method isvalidated through the noise reducing of the actual engine health signals.
In view of the problem that the process neural network (PNN) with continuousfunction inputs can not be trained using discrete samples directly, a discrete inputprocess neural network model (DPNN) is presented. DPNN takes discrete vectors asinputs and utlize convolution sum to realize the time aggregate operation. Thus,DPNN can avoid problems such as parameter selection and information loss duringprocedures of function fitting and expanding required by PNN with continuousfunction inputs. The engine health condition prediction results show that, DPNNperforms comparable to PNN with continuous function inputs while with betteroperability.
To overcome the difficulties of model optimization required by single globalmodels for engine health condition prediction, two ensemble prediction models with static and dynamic combining weights respectively based on the improvedAdaBoost.RT are proposed. The error function of AdaBoost.RT is improved, and anadaptive adjustment strategy is adopted to adjust the threshold during the trainingprocess. Then, DPNN, ANN and the extreme learning machine are utilized as weaklearners to construct ensemble models for engine health condition prediction. Forensemble models with dynamic combining weights, the weak learners are evaluatedusing the neighboring samples during the training process, and the dynamiccombining weights of the weak learners are generated according to their performanceon the neighboring samples of the testing sample in the training sample set. Since thelocal performance of the weak learners can be well mined, the ensemble models withdynamic combining weights are superior to their counterparts with static combiningweights. The prediction results of engine health conditions show that the ensemblemodels always perform better than single global models. And the performancerequirements for predictors can be reduced while being utilized as weak leaners.
According to the fact that the point prediction results are accompanied withuncertainty, the Bootstrap based prediction interval estimation method is employed toestimate the prediction intervals of the presented models, thus to achieve quantitativeassessment of the reliability and accuracy of the prediction results.
Based on such reseach, the “system for engine removal date prediction” isdeveloped according to the actual requirements of Air China. As a key subsystem ofthe “aircraft engine health management and maintenance decision support system”,the developed system acquires health condition information from its parent system toexecute engine removal date prediction, and the results are provided to themaintenance planning module for maintenance plan making, which implementsseamless integration from engine condition monitoring to maintenance decisionmaking.
引文
[1]左洪福,张海军,戎翔.基于比例风险模型的航空发动机视情维修决策[J].航空动力学报,2006,(4):716-721
[2]梁剑.基于成本优化的民用航空发动机视情维修决策研究[D].南京:南京航空航天大学博士学位论文.2004:1-5
[3]莫欣农. MRO2推动制造业向生产性服务转变[J].中国制造业信息化,2012,(18):18-20
[4]王建民. MRO2:制造业新的增长点[J].中国制造业信息化,2010,(10):13-15
[5]曾声奎, Michael G P,吴际.故障预测与健康管理(PHM)技术的现状与发展[J].航空学报,2005,(5):626-632
[6]张宝珍,曾天翔.先进的故障预测与状态管理技术[J].测控技术,2003,(11):4-6
[7]彭宇,刘大同,彭喜元.故障预测与健康管理技术综述[J].电子测量与仪器学报,2010,(1):1-9
[8] Scanff E, Feldman K L, Ghelam S, et al. Life Cycle Cost Impact of UsingPrognostic Health Management (PHM) for Helicopter Avionics[J].Microelectronics Reliability,2007,47(12):1857-1864
[9] Vichare N M, Pecht M G. Prognostics and Health Management of Electronics[J]. IEEE Transactions on Components and Packaging Technologies,2006,29(1):222-229
[10] Batzel T D, Swanson D C. Prognostic Health Management of Aircraft PowerGenerators [J]. IEEE Transactions on Aerospace and Electronic Systems,2009,45(2):473-482
[11] Byington C S, Watson M, Edwards D, et al. A Model-Based Approach toPrognostics and Health Management for Flight Control Actuators [C].Aerospace Conference,2004. Proceedings.2004IEEE, Big Sky, MT,2004.IEEE,2004:3551-3562
[12] De Oliveira Bizarria C, Yoneyama T. Prognostics and Health Monitoring foran Electro-Hydraulic Flight Control Actuator [C]. Aerospace conference,2009IEEE, Big Sky, MT,2009. IEEE,2009:1-9
[13] Millar R C. A Systems Engineering Approach to PHM For Military AircraftPropulsion Systems [C].2007IEEE Aerospace Conference, Big Sky, MT,2007. IEEE,2007:3727-3735
[14] Suarez E L, Duffy M J, Gamache R N, et al. Jet Engine Life PredictionSystems Integrated with Prognostics Health Management [C].2004IEEEAerospace Conference Proceedings, Big Sky, MT,2004. IEEE,2004:3596-3602
[15]钟诗胜.航空发动机MRO技术及其发展建议[J].国防制造技术,2010,(03):21-23
[16] Schwabacher M, Goebel K. A Survey of Artificial Intelligence forPrognostics [C]. Proceedings of2007AAAI Fall Symposium: AI forPrognostics, Arlington, VA, USA,2007.2007:107-114
[17] Hornik K, Stinchcombe M, White H. Multilayer Feedforward Networks areUniversal Approximators [J]. Neural Networks,1989,2(5):359-366
[18]何新贵,梁久祯.过程神经元网络的若干理论问题[J].中国工程科学,2000,(12):40-44
[19]陈果.用结构自适应神经网络预测航空发动机性能趋势[J].航空学报,2007,(03):535-539
[20]张营,左洪福,任淑红,等.一种航空发动机性能衰退预测的优化算法[J].中国机械工程,2011,(09):1009-1013
[21]何新贵,梁久祯,许少华.过程神经网络的训练及其应用[J].中国工程科学,2001,(04):31-35
[22] Freund Y, Schapire R E. A Desicion-Theoretic Generalization of On-LineLearning and an Application to Boosting [J]. Computational LearningTheory,1995,904:23-37
[23] Efron B. Bootstrap Methods: Another Look at the Jackknife [J]. The annalsof Statistics,1979,7(1):1-26
[24] Jaw L C, Friend R. ICEMS: a Platform for Advanced Condition-BasedHealth Management [C]. Aerospace Conference,2001, IEEE Proceedings,Big Sky, MT,2001. IEEE,2001:2909-2914
[25]梁旭,李行善,张磊,等.支持视情维修的故障预测技术研究[J].测控技术,2007,26(6):5-8
[26] Kalgren P W, Byington C S, Roemer M J. Defining PHM, A LexicalEvolution of Maintenance and Logistics [C]. Autotestcon,2006IEEE,2006.IEEE,2006:353-358
[27] Dragomir O E, Gouriveau R, Dragomir F, et al. Review of PrognosticProblem in Condition-Based Maintenance [C]. European Control Conference,Budapest, Hungary,2009.2009:6
[28] Sheppard J W, Kaufman M A, Wilmer T J. IEEE Standards for Prognosticsand Health Management [J]. Aerospace and Electronic Systems Magazine,IEEE,2009,24(9):34-41
[29] Verma R, Ganguli R. Denoising Jet Engine Gas Path Measurements UsingNonlinear Filters [J]. IEEE/ASME Transactions on Mechatronics,2005,10(4):461-464
[30] Ganguli R. Noise and Outlier Removal From Jet Engine Health SignalsUsing Weighted Fir Median Hybrid Filters [J]. Mechanical Systems andSignal Processing,2002,16(6):967-978
[31] Ganguli R, Dan B. Trend Shift Detection in Jet Engine Gas PathMeasurements Using Cascaded Recursive Median Filter With Gradient andLaplacian Edge Detector [J]. Journal of Engineering for Gas Turbines andPower,2004,126(1):55
[32]栾圣罡.基于气路参数样本的航空发动机状态监视方法与系统研究[D].哈尔滨工业大学博士学位论文.2008:16-33
[33] Ganguli R. Data Rectification and Detection of Trend Shifts in Jet EnginePath Measurements Using Median Filters and Fuzzy Logic [J]. Journal ofEngineering for Gas Turbines and Power,2002,124(4):809
[34] Uday P, Ganguli R. Jet Engine Health Signal Denoising Using OptimallyWeighted Recursive Median Filters [J]. Journal of Engineering for GasTurbines and Power,2010,132(4):041601
[35]马建仓,刘小龙,陈静.二代小波降噪与盲分离结合应用于航空发动机振动信号分析[J].机械科学与技术,2010,(01):7-11
[36]雷衍斌,李舜酩,门秀花,等.基于自相关降噪的混叠转子振动信号分离[J].振动与冲击,2011,(01):218-222
[37]史永胜,岳俊杰,宋云雪.非线性去噪对发动机排气温度预测精度影响[J].计算机工程与应用,2012,48(25):212-216
[38]吕永乐,郎荣玲.基于奇异值分解的飞行数据降噪方法[J].计算机工程,2010,36(03):260-262
[39]朱玉斌,方中祥,侯朝焕,等.基于部件参数估计的航空发动机故障预测技术研究[J].测控技术,2012,31(01):128-131
[40] Kobayashi, Takahisa, Simon D L. Integration of On-Line and Off-LineDiagnostic Algorithms for Aircraft Engine Health Management [J].Transactions-Asme Journal of Engineering for Gas Turbines and Power,2007,129(4):986
[41] Marinai L, Singh R, Curnock B, et al. Detection and Prediction of thePerformance Deterioration of a Turbofan Engine [C]. Proceedings of theInternational Gas Turbine Congress Tokyo,2003.2003:1-7
[42]李晓白,崔秀伶,郎荣玲.航空发动机性能参数预测方法[J].北京航空航天大学学报,2008,(03):253-256
[43]干敏梁,左洪福,杨忠,等.时序建模方法在滑油光谱分析中的应用[J].光谱学与光谱分析,2000,(01):64-67
[44]胡金海,谢寿生.基于AR模型对滑油中金属元素含量的预测[J].燃气涡轮试验与研究,2003,(01):32-36
[45]宋云雪,张科星,史永胜.基于多元线性回归的发动机性能参数预测[J].航空动力学报,2009,(02):427-431
[46]李本威,张赟,孙涛.基于免疫粒子群算法的滑油屑末支持向量机预测模型设计[J].航空动力学报,2009,(07):1639-1643
[47]王旭辉,黄圣国,曹力,等.基于LS-SVM的航空发动机气路参数趋势在线预测[J].吉林大学学报(工学版),2008,(01):239-244
[48] Xu K. Application of Neural Networks in Forecasting Engine SystemsReliability [J]. Applied Soft Computing,2003,2(4):255-268
[49]魏海坤,徐嗣鑫,宋文忠.神经网络的泛化理论和泛化方法[J].自动化学报,2001,(06):806-815
[50]许少华,何新贵.基于函数正交基展开的过程神经网络学习算法[J].计算机学报,2004,(05):645-650
[51]李盼池,王海英.基于样条插值函数的离散过程神经网络训练[J].计算机应用研究,2011,28(01):75-77
[52]李盼池,王海英,戴庆,肖红.量子过程神经网络模型算法及应用[J].物理学报,2012,16:24-32
[53]何新贵,许少华.一类反馈过程神经元网络模型及其学习算法[J].自动化学报,2004,(06):801-806
[54]许少华,何新贵.多聚合过程神经元网络及其学习算法研究[J].计算机学报,2007,(01):48-56
[55]许少华,何新贵,李盼池.自组织过程神经网络及其应用研究[J].计算机研究与发展,2003,(11):1612-1615
[56]许少华,李盼池,何新贵.一种概率过程神经元网络模型及分类算法[J].智能系统学报,2009,4(18):283-287
[57]李洋.小波过程神经网络相关理论及其应用研究[D].哈尔滨工业大学博士学位论文.2008:14-15
[58] Zhong S S, Li Y, Ding G, et al. Continuous Wavelet Process Neural Networkand its Application [J]. Neural network world,2007,16:483-495
[59]丁刚.面向状态预报的过程神经网络模型及其应用研究[D].哈尔滨:哈尔滨工业大学博士学位论文.2006:96-101
[60] Zhong S S, Ding G, Su D Z. Parallel Feedforward Process Neural Networkwith Time-Varying Input and Output Functions[J]. Advances in NeuralNetworks,2005,3496:103-124
[61]丁刚,付旭云,钟诗胜.基于过程神经网络的航空发动机性能参数预测[J].计算机集成制造系统,2011,17(01):198-207
[62]丁刚,徐敏强,侯立国.基于过程神经网络的航空发动机排气温度预测[J].航空动力学报,2009,24(05):1035-1039
[63]钟诗胜,栾圣罡,丁刚.基于过程神经网络与气动热力参数的航空发动机状态监视[J].南京理工大学学报(自然科学版),2006,(05):533-536+541
[64]钟诗胜,李洋.基于小波过程神经网络的飞机发动机状态监视[J].航空学报,2007,(01):68-71
[65]孙智源.基于过程神经网络集成的航空发动机性能衰退预测[D].哈尔滨工业大学硕士学位论文.2010:14-16
[66] Ding G, Zhong S S. Approximation Capability Analysis of Parallel ProcessNeural Network with Application to Aircraft Engine Health ConditionMonitoring [J]. Advances in Neural Networks,2007,4493:66-72
[67]付旭云,丁刚,钟诗胜,等.基于双隐层径向基过程神经网络的润滑油金属含量预测[J].润滑与密封,2009,(02):17-20+51
[68]金向阳,林琳,钟诗胜,等.航空发动机振动趋势预测的过程神经网络法[J].振动.测试与诊断,2011,31(03):331-334+396-397
[69]鲁峰,黄金泉.航空发动机部件性能参数融合预测[J].航空学报,2009,(10):1795-1800
[70]陈果,杨虞微.航空发动机复杂磨损趋势的神经网络多变量预测模型[J].中国机械工程,2007,(01):70-74
[71]吕永乐,郎荣玲,路辉,等.航空发动机性能参数联合RBFPN和FAR预测[J].北京航空航天大学学报,2010,(02):131-134+149
[72] Pecht M, Jaai R. A Prognostics and Health Management Roadmap forInformation and Electronics-Rich Systems [J]. Microelectronics Reliability,2010,50(3):317-323
[73] Heng A, Zhang S, Tan A C C, et al. Rotating Machinery Prognostics: State ofthe Art, Challenges and Opportunities [J]. Mechanical Systems and SignalProcessing,2009,23(3):724-739
[74] Heng A, Tan A C C, Mathew J, et al. Intelligent Condition-based Predictionof Machinery Reliability [J]. Mechanical Systems and Signal Processing,2009,23(5):1600-1614
[75] Tian Z G. A Neural Network Approach for Remaining Useful Life PredictionUtilizing Both Failure and Suspension Data [C]. Reliability andMaintainability Symposium (RAMS),2010Proceedings-Annual, San Jose,CA,2010. IEEE,2010:1-6
[76] Tian Z. An Artificial Neural Network Approach for Remaining Useful LifePrediction of Equipments Subject to Condition Monitoring [C].8thInternational Conference on Reliability, Maintainability and Safety,2009(ICRMS2009), Chengdu, China,2009. IEEE,2009:143-148
[77] zg ren Y, etinkaya S, Sar demir S, et al. Predictive Modeling OfPerformance of a Helium Charged Stirling Engine Using an Artificial NeuralNetwork [J]. Energy Conversion and Management,2013,67(0):357-368
[78] Yuliang D, Yujiong G, Kun Y, et al. Applying PCA to Establish ArtificialNeural Network for Condition Prediction on Equipment in Power Plant [C].Fifth World Congress on Intelligent Control and Automation,2004(WCICA2004),2004. IEEE,2004:1715-1719
[79] Yu P, Hong W, Jianmin W, et al. A Modified Echo State Network basedRemaining Useful Life Estimation Approach [C].2012IEEE Conference onPrognostics and Health Management (PHM) Denver, CO,2012. IEEE,2012:1-7
[80] Byington C S, Watson M, Edwards D. Data-driven Neural NetworkMethodology to Remaining Life Predictions for Aircraft ActuatorComponents [C].2004IEEE Aerospace Conference,2004. Proceedings,2004. IEEE,2004:3581-3589
[81] Roemer M J, Byington C S, Kacprzynski G J, et al. An Overview of SelectedPrognostic Technologies with Application to Engine Health Management [C].ASME Turbo Expo2006, Barcelona, Spain,2006. ASME,2006:707-715
[82]丁刚,钟诗胜.基于过程神经网络的热平衡温度预测研究[J].宇航学报,2006,(03):489-492+545
[83]杨海,程伟,楚丽妍.基于过程神经网络算法的航天器非平稳随机振动时频分析[J].振动与冲击,2008,(01):12-15+29+179
[84] Breiman L. Bagging Predictors [J]. Machine Learning,1996,24(2):123-140
[85] Dietterich T. Ensemble Methods in Machine Learning [J]. MultipleClassifier Systems,2000,1857:1-15
[86] Soelaiman R, Martoyo A, Purwananto Y, et al. Implementation of RecurrentNeural Network and Boosting Method for Time-Series Forecasting [C].ICICI-BME:2009International Conference on Instrumentation,Communication, Information Technology, and Biomedical Engineering,Bandung,2009. IEEE,2009:55-62
[87] Li Y, Fujimoto R M, Hunter M P. Online Travel Time Prediction Based onBoosting [C].200912th International IEEE Conference on IntelligentTransportation Systems (ITSC2009), St. Louis, MO,2009. IEEE,2009:759-764
[88] Li H, Zeng M, Lu M, et al. Adaboosting-based Dynamic WeightedCombination of Software Reliability Growth Models[J]. Quality andReliability Engineering International,2012,28(1):67-84
[89] Hao X, Meng F, Zhou Y. An Improved Adaboost.R Algorithm and ItsApplication in Mining Safety Monitoring [C]. Third InternationalSymposium on Intelligent Information Technology Application,2009(IITA2009), Nanchang, China,2009. IEEE,2009:287-290
[90] Xu R Z, Geng X F, Zhou F Y. A Short Term Load Forecasting Based onBagging-ELM Algorithm [C]. Proceedings of the2012InternationalConference on Information Technology and Software Engineering, Beijing,2013. Springer Berlin Heidelberg:507-514
[91] Gao F, Kou P, Gao L, et al. Boosting Regression Methods Based on aGeometric Conversion Approach: Using SVMs Base Learners [J].Neurocomputing,2013,113:67-87
[92]赵学智,叶邦彦. SVD和小波变换的信号处理效果相似性及其机理分析[J].电子学报,2008,36(08):1582-1589
[93]张波,李健君.基于Hankel矩阵与奇异值分解(SVD)的滤波方法以及在飞机颤振试验数据预处理中的应用[J].振动与冲击,2009,28(02):162-166+208
[94] Jha S K, Yadava R D S. Denoising by Singular Value Decomposition and ItsApplication to Electronic Nose Data Processing [J]. Sensors Journal, IEEE,2011,11(1):35-44
[95] Zhao X Z, Ye B Y. Selection of Effective Singular Values Using DifferenceSpectrum and Its Application to Fault Diagnosis of Headstock [J].Mechanical Systems and Signal Processing,2011,25(5):1617-1631
[96] Huang N E, Shen Z, Long S R, et al. The Empirical Mode Decompositionand the Hilbert Spectrum for Nonlinear and Non-Stationary Time SeriesAnalysis [J]. Proceedings of the Royal Society A: Mathematical, Physicaland Engineering Sciences,1998,454(1971):903-995
[97] Flandrin P, Gonc alves P, Rilling G. Detrending and Denoising withEmpirical Mode Decompositions [C]. Proceedings of the12th EuropeanSignal Processing Conference (EUSIPCO'04), Vienna, Austria,2004.2004:1581-1584
[98] Wu Z, Huang N E, Long S R, et al. On the Trend, Detrending, and Variabilityof Nonlinear and Nonstationary Time Series [J]. Proc Natl Acad Sci U S A,2007,104(38):14889-14894
[99] Rilling G, Flandrin P, Gon alvés P. On Empirical Mode Decomposition andits Algorithms [C]. IEEE-EURASIP Workshop on Nonlinear Signal andImage Processing (NSIP’03), Grado, Italy,2003.2003:8-11
[100]丁刚,钟诗胜,栾圣罡.航空发动机滑油金属含量预测及其控制研究[J].润滑与密封,2006,(09):52-54
[101]欧阳楷，邹睿，刘卫芳.基于生物的神经网络的理论框架──神经元模型[J].北京生物医学工程,1997,(02):93-101
[102] Hagan M T, Menhaj M B. Training Feedforward Networks with theMarquardt Algorithm [J]. IEEE Transactions on Neural Networks,1994,5(6):989-993
[103] Mackey M C, Glass L. Oscillation and Chaos in Physiological ControlSystems [J]. Science,1977,197(4300):287-289
[104] Schapire R E. The Strength of Weak Learnability [J]. Machine learning,1990,5(2):197-227
[105] Polikar R. Ensemble Based Systems in Decision Making [J]. Circuits andSystems Magazine, IEEE,2006,6(3):21-45
[106] Solomatine D P, Shrestha D L. AdaBoost.RT: a Boosting Algorithm forRegression Problems [C].2004IEEE International Joint Conference onNeural Networks,2004. IEEE,2004:1163-1168
[107] Geman S, Bienenstock E, Doursat R. Neural Networks and the Bias/VarianceDilemma [J]. Neural computation,1992,4(1):1-58
[108] Ueda N, Nakano R. Generalization Error of Ensemble Estimators [C]. IEEEInternational Conference on Neural Networks,1996, Washington, DC,1996.IEEE,1996:90-95
[109] Zhou Z H. Ensemble Methods: Foundations and Algorithms [M]. Chapman&Hall,2012:99-102
[110] Solomatine D P. Local and Hybrid Learning Models in Forecasting NaturalPhenomena [C]. Proceedings of the International Joint Conference on NeuralNetworks (IJCNN),2005. IEEE,2005:1716-1721
[111] Yoav F, Schapire R E. Experiments with a New Boosting Algorithm [C].Machine Learning: Proceedings of the Thirteenth Conference,1996. MorganKaufmann Publishers, Inc,1996:148-156
[112] Drucker H. Improving Regressors using Boosting Techniques [C]. ICML '97Proceedings of the Fourteenth International Conference on MachineLearning, San Francisco, CA, USA,1997. San Francisco, CA, USA, MorganKaufmann Publishers Inc,1997:107-115
[113] Yang Y H, Lin Y C, Su Y F, et al. Music Emotion Classification: ARegression Approach [C].2007IEEE International Conference onMultimedia and Expo, Beijing, China,2007. IEEE,2007:208-211
[114] Tian H X, Mao Z Z. An Ensemble ELM Based on Modified AdaBoost.RTAlgorithm for Predicting the Temperature of Molten Steel in Ladle Furnace[J]. IEEE Transactions on Automation Science and Engineering,2010,7(1):73-80
[115] Costa E O, Pozo A, Vergilio S R. Using Boosting Techniques to ImproveSoftware Reliability Models Based on Genetic Programming [C].ICTAI-2006: Eighteenth International Conference on Tools with ArtificialIntelligence, Proceedings, Arlington, VA,2006. IEEE,2006:643-650
[116] Bar A, Rokach L, Shani G, et al. Improving Simple Collaborative FilteringModels Using Ensemble Methods [C].11th International Workshop, MCS2013, Nanjing, China,2013. Springer Berlin Heidelberg:1-12
[117] Shrestha D L, Solomatine D P. Experiments with AdaBoost.RT, an ImprovedBoosting Scheme for Regression [J]. Neural Comput,2006,18(7):1678-1710
[118] Moreiraa J M, Soaresb C, Jorgeb A M, et al. Ensemble Approaches forRegression: A Survey [J]. ACM Computing Surveys,2007,45(1):
[119] Yeung D S, Chan P P. A Novel Dynamic Fusion Method Using LocalizedGeneralization Error Model [C]. IEEE International Conference on Systems,Man and Cybernetics, San Antonio, TX,2009. IEEE,2009:623-628
[120] Cao D S, Xu Q S, Liang Y Z, et al. The Boosting: A New Idea of BuildingModels [J]. Chemometrics and Intelligent Laboratory Systems,2010,100(1):1-11
[121] Niall R, Patterson D. A Fusion of Stacking with Dynamic Integration [C].20th International Joint Conference on Artificial Intelligence, Hyderabad,India,2007. Morgan Kaufmann Publishers Inc,2007:2844-2849
[122] Puuronen S, Terziyan V, Tsymbal A. A Dynamic Integration Algorithm for anEnsemble of Classifiers [C].11th International Symposium, Foundations ofIntelligent Systems Warsaw, Poland,1999. Springer Berlin Heidelberg,1999:592-600
[123] Alexey T, Puuronen S. Bagging and Boosting with Dynamic Integration ofClassifiers [C].4th European Conference, Principles of Data Mining andKnowledge Discovery Lyon, France,2000. Springer Berlin Heidelberg,2000:116-125
[124] Jo o M M, Jorge A M, Soares C, et al. Ensemble Learning: A Study onDifferent Variants of the Dynamic Selection Approach [C].6th InternationalConference, Machine Learning and Data Mining in Pattern Recognition,Leipzig, Germany,2009. Springer Berlin Heidelberg,2009:191-205
[125] Niall R, Patterson D, Anand S, et al. Dynamic Integration of RegressionModels [C]. Proceedings of the5th International Multiple Classifier SystemsWorkshop, Cagliari, Italy,2004. Springer Berlin Heidelberg,2004:164-173
[126] Alexey T, Puuronen S, Terziyan V. A Technique for Advanced DynamicIntegration of Multiple Classifiers [C]. Proceedings of the FinnishConference on Artificial Intelligence,1998.1998:71-79
[127] Huang G B, Zhu Q Y, Siew C K. Extreme Learning Machine: Theory andApplications [J]. Neurocomputing,2006,70(1-3):489-501
[128] Fukunaga K, Narendra P M. A Branch and Bound Algorithm for ComputingK-Nearest Neighbors [J]. IEEE Transactions on Computers,1975,100(7):750-753
[129] Ko A H, Sabourin R, Britto Jr A S. From Dynamic Classifier Selection toDynamic Ensemble Selection [J]. Pattern Recognition,2008,41(5):1718-1731
[130] Woods K, Kegelmeyer W P, Bowyer K. Combination of Multiple ClassifiersUsing Local Accuracy Estimates [J]. IEEE Transactions on Pattern Analysisand Machine Intelligence,1997,9(4):405-410
[131] Yeung D S, Ng W W, Wang D, et al. Localized Generalization Error Modeland Its Application to Architecture Selection for Radial Basis FunctionNeural Network [J]. IEEE Trans Neural Netw,2007,18(5):1294-1305
[132] Zhang C X, Zhang J S. A Local Boosting Algorithm for SolvingClassification Problems [J]. Computational Statistics&Data Analysis,2008,52(4):1928-1941
[133] Huang G B, Wang D H, Lan Y. Extreme Learning Machines: A Survey [J].International Journal of Machine Learning and Cybernetics,2011,2(2):107-122
[134] Huang G B, Zhou H M, Ding X J, et al. Extreme Learning Machine forRegression And Multiclass Classification [J]. IEEE Transactions on Systems,Man, and Cybernetics, Part B: Cybernetics,2012,42(2):513-529
[135] http://www.stat.wisc.edu/~reinsel/bjr-data/
[136] Khosravi A, Nahavandi S, Creighton D, et al. Comprehensive Review ofNeural Network-Based Prediction Intervals and New Advances[J]. IEEETrans Neural Netw,2011,22(9):1341-1356
[137] Heskes T. Practical Confidence and Prediction Intervals [J]. Advances inneural information processing systems1997:176-182
[138] Khosravi A, Nahavandi S, Creighton D. Construction of Optimal PredictionIntervals for Load Forecasting Problems [J]. IEEE Transactions on PowerSystems,2010,25(3):1496-1503
[139] Zio E. A Study of the Bootstrap Method for Estimating the Accuracy ofArtificial Neural Networks in Predicting Nuclear Transient Processes [J].IEEE Transactions on Nuclear Science,2006,53(3):1460-1478
[140] Oleng N O, Gribok A, Reifman J. Error Bounds for Data-Driven Models ofDynamical Systems [J]. Comput Biol Med,2007,37(5):670-679
[141] De Veaux R D, Schumi J, Schweinsberg J, et al. Prediction Intervals forNeural Networks Via Nonlinear Regression[J]. Technometrics,1998,40(4):273-282
[142] MacKay D J. The Evidence Framework Applied to Classification Networks[J]. Neural computation1992,4(5):720-736
[143] Nix D A, Weigend A S. Estimating the Mean and Variance of the TargetProbability Distribution [C].1994. IEEE World Congress on ComputationalIntelligence,1994IEEE International Conference on Neural Networks,Orlando, FL,1994. IEEE,1994:55-60
[144] Khosravi A, Nahavandi S, Creighton D, et al. Optimizing The Quality ofBootstrap-based Prediction Intervals [C].2011International JointConference on Neural Networks (Ijcnn), San Jose, CA,2011. IEEE,2011:3072-3078
[145] Riedmiller M, Braun H. A Direct Adaptive Method for FasterBackpropagation Learning: the Rprop Algorithm [C]. IEEE InternationalConference on Neural Networks, San Francisco, CA,1993. IEEE,1993:586-591
[146] Ahmad F, Isa N A M, Osman M K, et al. Performance Comparison ofGradient Descent and Genetic Algorithm Based Artificial Neural NetworksTraining [C]. IEEE10th International Conference on Intelligent SystemsDesign and Applications (ISDA), Cairo,2010. IEEE,2010:604-609
[147] Anastasiadis A D, Magoulas G D, Vrahatis M N. New Globally ConvergentTraining Scheme Based on the Resilient Propagation Algorithm [J].Neurocomputing,2005,64:253-270
[148] Santosh T V, Vinod G, Saraf R K, et al. Application of Artificial NeuralNetworks to Nuclear Power Plant Transient Diagnosis[J]. ReliabilityEngineering&System Safety,2007,92(10):1468-1472
[149] Nada R H, Habashy D M, El-Salam F A, et al. Modeling of Aging Processfor Supersaturated Solution Treated of Al–3wt%Mg Alloy [J]. MaterialsScience and Engineering: A,2013,567(0):80-83
[150] Gebraeel N, Elwany A, Jing P. Residual Life Predictions in the Absence ofPrior Degradation Knowledge [J]. IEEE Transactions on Reliability,2009,58(1):106-117
[2]梁剑.基于成本优化的民用航空发动机视情维修决策研究[D].南京:南京航空航天大学博士学位论文.2004:1-5
[3]莫欣农. MRO2推动制造业向生产性服务转变[J].中国制造业信息化,2012,(18):18-20
[4]王建民. MRO2:制造业新的增长点[J].中国制造业信息化,2010,(10):13-15
[5]曾声奎, Michael G P,吴际.故障预测与健康管理(PHM)技术的现状与发展[J].航空学报,2005,(5):626-632
[6]张宝珍,曾天翔.先进的故障预测与状态管理技术[J].测控技术,2003,(11):4-6
[7]彭宇,刘大同,彭喜元.故障预测与健康管理技术综述[J].电子测量与仪器学报,2010,(1):1-9
[8] Scanff E, Feldman K L, Ghelam S, et al. Life Cycle Cost Impact of UsingPrognostic Health Management (PHM) for Helicopter Avionics[J].Microelectronics Reliability,2007,47(12):1857-1864
[9] Vichare N M, Pecht M G. Prognostics and Health Management of Electronics[J]. IEEE Transactions on Components and Packaging Technologies,2006,29(1):222-229
[10] Batzel T D, Swanson D C. Prognostic Health Management of Aircraft PowerGenerators [J]. IEEE Transactions on Aerospace and Electronic Systems,2009,45(2):473-482
[11] Byington C S, Watson M, Edwards D, et al. A Model-Based Approach toPrognostics and Health Management for Flight Control Actuators [C].Aerospace Conference,2004. Proceedings.2004IEEE, Big Sky, MT,2004.IEEE,2004:3551-3562
[12] De Oliveira Bizarria C, Yoneyama T. Prognostics and Health Monitoring foran Electro-Hydraulic Flight Control Actuator [C]. Aerospace conference,2009IEEE, Big Sky, MT,2009. IEEE,2009:1-9
[13] Millar R C. A Systems Engineering Approach to PHM For Military AircraftPropulsion Systems [C].2007IEEE Aerospace Conference, Big Sky, MT,2007. IEEE,2007:3727-3735
[14] Suarez E L, Duffy M J, Gamache R N, et al. Jet Engine Life PredictionSystems Integrated with Prognostics Health Management [C].2004IEEEAerospace Conference Proceedings, Big Sky, MT,2004. IEEE,2004:3596-3602
[15]钟诗胜.航空发动机MRO技术及其发展建议[J].国防制造技术,2010,(03):21-23
[16] Schwabacher M, Goebel K. A Survey of Artificial Intelligence forPrognostics [C]. Proceedings of2007AAAI Fall Symposium: AI forPrognostics, Arlington, VA, USA,2007.2007:107-114
[17] Hornik K, Stinchcombe M, White H. Multilayer Feedforward Networks areUniversal Approximators [J]. Neural Networks,1989,2(5):359-366
[18]何新贵,梁久祯.过程神经元网络的若干理论问题[J].中国工程科学,2000,(12):40-44
[19]陈果.用结构自适应神经网络预测航空发动机性能趋势[J].航空学报,2007,(03):535-539
[20]张营,左洪福,任淑红,等.一种航空发动机性能衰退预测的优化算法[J].中国机械工程,2011,(09):1009-1013
[21]何新贵,梁久祯,许少华.过程神经网络的训练及其应用[J].中国工程科学,2001,(04):31-35
[22] Freund Y, Schapire R E. A Desicion-Theoretic Generalization of On-LineLearning and an Application to Boosting [J]. Computational LearningTheory,1995,904:23-37
[23] Efron B. Bootstrap Methods: Another Look at the Jackknife [J]. The annalsof Statistics,1979,7(1):1-26
[24] Jaw L C, Friend R. ICEMS: a Platform for Advanced Condition-BasedHealth Management [C]. Aerospace Conference,2001, IEEE Proceedings,Big Sky, MT,2001. IEEE,2001:2909-2914
[25]梁旭,李行善,张磊,等.支持视情维修的故障预测技术研究[J].测控技术,2007,26(6):5-8
[26] Kalgren P W, Byington C S, Roemer M J. Defining PHM, A LexicalEvolution of Maintenance and Logistics [C]. Autotestcon,2006IEEE,2006.IEEE,2006:353-358
[27] Dragomir O E, Gouriveau R, Dragomir F, et al. Review of PrognosticProblem in Condition-Based Maintenance [C]. European Control Conference,Budapest, Hungary,2009.2009:6
[28] Sheppard J W, Kaufman M A, Wilmer T J. IEEE Standards for Prognosticsand Health Management [J]. Aerospace and Electronic Systems Magazine,IEEE,2009,24(9):34-41
[29] Verma R, Ganguli R. Denoising Jet Engine Gas Path Measurements UsingNonlinear Filters [J]. IEEE/ASME Transactions on Mechatronics,2005,10(4):461-464
[30] Ganguli R. Noise and Outlier Removal From Jet Engine Health SignalsUsing Weighted Fir Median Hybrid Filters [J]. Mechanical Systems andSignal Processing,2002,16(6):967-978
[31] Ganguli R, Dan B. Trend Shift Detection in Jet Engine Gas PathMeasurements Using Cascaded Recursive Median Filter With Gradient andLaplacian Edge Detector [J]. Journal of Engineering for Gas Turbines andPower,2004,126(1):55
[32]栾圣罡.基于气路参数样本的航空发动机状态监视方法与系统研究[D].哈尔滨工业大学博士学位论文.2008:16-33
[33] Ganguli R. Data Rectification and Detection of Trend Shifts in Jet EnginePath Measurements Using Median Filters and Fuzzy Logic [J]. Journal ofEngineering for Gas Turbines and Power,2002,124(4):809
[34] Uday P, Ganguli R. Jet Engine Health Signal Denoising Using OptimallyWeighted Recursive Median Filters [J]. Journal of Engineering for GasTurbines and Power,2010,132(4):041601
[35]马建仓,刘小龙,陈静.二代小波降噪与盲分离结合应用于航空发动机振动信号分析[J].机械科学与技术,2010,(01):7-11
[36]雷衍斌,李舜酩,门秀花,等.基于自相关降噪的混叠转子振动信号分离[J].振动与冲击,2011,(01):218-222
[37]史永胜,岳俊杰,宋云雪.非线性去噪对发动机排气温度预测精度影响[J].计算机工程与应用,2012,48(25):212-216
[38]吕永乐,郎荣玲.基于奇异值分解的飞行数据降噪方法[J].计算机工程,2010,36(03):260-262
[39]朱玉斌,方中祥,侯朝焕,等.基于部件参数估计的航空发动机故障预测技术研究[J].测控技术,2012,31(01):128-131
[40] Kobayashi, Takahisa, Simon D L. Integration of On-Line and Off-LineDiagnostic Algorithms for Aircraft Engine Health Management [J].Transactions-Asme Journal of Engineering for Gas Turbines and Power,2007,129(4):986
[41] Marinai L, Singh R, Curnock B, et al. Detection and Prediction of thePerformance Deterioration of a Turbofan Engine [C]. Proceedings of theInternational Gas Turbine Congress Tokyo,2003.2003:1-7
[42]李晓白,崔秀伶,郎荣玲.航空发动机性能参数预测方法[J].北京航空航天大学学报,2008,(03):253-256
[43]干敏梁,左洪福,杨忠,等.时序建模方法在滑油光谱分析中的应用[J].光谱学与光谱分析,2000,(01):64-67
[44]胡金海,谢寿生.基于AR模型对滑油中金属元素含量的预测[J].燃气涡轮试验与研究,2003,(01):32-36
[45]宋云雪,张科星,史永胜.基于多元线性回归的发动机性能参数预测[J].航空动力学报,2009,(02):427-431
[46]李本威,张赟,孙涛.基于免疫粒子群算法的滑油屑末支持向量机预测模型设计[J].航空动力学报,2009,(07):1639-1643
[47]王旭辉,黄圣国,曹力,等.基于LS-SVM的航空发动机气路参数趋势在线预测[J].吉林大学学报(工学版),2008,(01):239-244
[48] Xu K. Application of Neural Networks in Forecasting Engine SystemsReliability [J]. Applied Soft Computing,2003,2(4):255-268
[49]魏海坤,徐嗣鑫,宋文忠.神经网络的泛化理论和泛化方法[J].自动化学报,2001,(06):806-815
[50]许少华,何新贵.基于函数正交基展开的过程神经网络学习算法[J].计算机学报,2004,(05):645-650
[51]李盼池,王海英.基于样条插值函数的离散过程神经网络训练[J].计算机应用研究,2011,28(01):75-77
[52]李盼池,王海英,戴庆,肖红.量子过程神经网络模型算法及应用[J].物理学报,2012,16:24-32
[53]何新贵,许少华.一类反馈过程神经元网络模型及其学习算法[J].自动化学报,2004,(06):801-806
[54]许少华,何新贵.多聚合过程神经元网络及其学习算法研究[J].计算机学报,2007,(01):48-56
[55]许少华,何新贵,李盼池.自组织过程神经网络及其应用研究[J].计算机研究与发展,2003,(11):1612-1615
[56]许少华,李盼池,何新贵.一种概率过程神经元网络模型及分类算法[J].智能系统学报,2009,4(18):283-287
[57]李洋.小波过程神经网络相关理论及其应用研究[D].哈尔滨工业大学博士学位论文.2008:14-15
[58] Zhong S S, Li Y, Ding G, et al. Continuous Wavelet Process Neural Networkand its Application [J]. Neural network world,2007,16:483-495
[59]丁刚.面向状态预报的过程神经网络模型及其应用研究[D].哈尔滨:哈尔滨工业大学博士学位论文.2006:96-101
[60] Zhong S S, Ding G, Su D Z. Parallel Feedforward Process Neural Networkwith Time-Varying Input and Output Functions[J]. Advances in NeuralNetworks,2005,3496:103-124
[61]丁刚,付旭云,钟诗胜.基于过程神经网络的航空发动机性能参数预测[J].计算机集成制造系统,2011,17(01):198-207
[62]丁刚,徐敏强,侯立国.基于过程神经网络的航空发动机排气温度预测[J].航空动力学报,2009,24(05):1035-1039
[63]钟诗胜,栾圣罡,丁刚.基于过程神经网络与气动热力参数的航空发动机状态监视[J].南京理工大学学报(自然科学版),2006,(05):533-536+541
[64]钟诗胜,李洋.基于小波过程神经网络的飞机发动机状态监视[J].航空学报,2007,(01):68-71
[65]孙智源.基于过程神经网络集成的航空发动机性能衰退预测[D].哈尔滨工业大学硕士学位论文.2010:14-16
[66] Ding G, Zhong S S. Approximation Capability Analysis of Parallel ProcessNeural Network with Application to Aircraft Engine Health ConditionMonitoring [J]. Advances in Neural Networks,2007,4493:66-72
[67]付旭云,丁刚,钟诗胜,等.基于双隐层径向基过程神经网络的润滑油金属含量预测[J].润滑与密封,2009,(02):17-20+51
[68]金向阳,林琳,钟诗胜,等.航空发动机振动趋势预测的过程神经网络法[J].振动.测试与诊断,2011,31(03):331-334+396-397
[69]鲁峰,黄金泉.航空发动机部件性能参数融合预测[J].航空学报,2009,(10):1795-1800
[70]陈果,杨虞微.航空发动机复杂磨损趋势的神经网络多变量预测模型[J].中国机械工程,2007,(01):70-74
[71]吕永乐,郎荣玲,路辉,等.航空发动机性能参数联合RBFPN和FAR预测[J].北京航空航天大学学报,2010,(02):131-134+149
[72] Pecht M, Jaai R. A Prognostics and Health Management Roadmap forInformation and Electronics-Rich Systems [J]. Microelectronics Reliability,2010,50(3):317-323
[73] Heng A, Zhang S, Tan A C C, et al. Rotating Machinery Prognostics: State ofthe Art, Challenges and Opportunities [J]. Mechanical Systems and SignalProcessing,2009,23(3):724-739
[74] Heng A, Tan A C C, Mathew J, et al. Intelligent Condition-based Predictionof Machinery Reliability [J]. Mechanical Systems and Signal Processing,2009,23(5):1600-1614
[75] Tian Z G. A Neural Network Approach for Remaining Useful Life PredictionUtilizing Both Failure and Suspension Data [C]. Reliability andMaintainability Symposium (RAMS),2010Proceedings-Annual, San Jose,CA,2010. IEEE,2010:1-6
[76] Tian Z. An Artificial Neural Network Approach for Remaining Useful LifePrediction of Equipments Subject to Condition Monitoring [C].8thInternational Conference on Reliability, Maintainability and Safety,2009(ICRMS2009), Chengdu, China,2009. IEEE,2009:143-148
[77] zg ren Y, etinkaya S, Sar demir S, et al. Predictive Modeling OfPerformance of a Helium Charged Stirling Engine Using an Artificial NeuralNetwork [J]. Energy Conversion and Management,2013,67(0):357-368
[78] Yuliang D, Yujiong G, Kun Y, et al. Applying PCA to Establish ArtificialNeural Network for Condition Prediction on Equipment in Power Plant [C].Fifth World Congress on Intelligent Control and Automation,2004(WCICA2004),2004. IEEE,2004:1715-1719
[79] Yu P, Hong W, Jianmin W, et al. A Modified Echo State Network basedRemaining Useful Life Estimation Approach [C].2012IEEE Conference onPrognostics and Health Management (PHM) Denver, CO,2012. IEEE,2012:1-7
[80] Byington C S, Watson M, Edwards D. Data-driven Neural NetworkMethodology to Remaining Life Predictions for Aircraft ActuatorComponents [C].2004IEEE Aerospace Conference,2004. Proceedings,2004. IEEE,2004:3581-3589
[81] Roemer M J, Byington C S, Kacprzynski G J, et al. An Overview of SelectedPrognostic Technologies with Application to Engine Health Management [C].ASME Turbo Expo2006, Barcelona, Spain,2006. ASME,2006:707-715
[82]丁刚,钟诗胜.基于过程神经网络的热平衡温度预测研究[J].宇航学报,2006,(03):489-492+545
[83]杨海,程伟,楚丽妍.基于过程神经网络算法的航天器非平稳随机振动时频分析[J].振动与冲击,2008,(01):12-15+29+179
[84] Breiman L. Bagging Predictors [J]. Machine Learning,1996,24(2):123-140
[85] Dietterich T. Ensemble Methods in Machine Learning [J]. MultipleClassifier Systems,2000,1857:1-15
[86] Soelaiman R, Martoyo A, Purwananto Y, et al. Implementation of RecurrentNeural Network and Boosting Method for Time-Series Forecasting [C].ICICI-BME:2009International Conference on Instrumentation,Communication, Information Technology, and Biomedical Engineering,Bandung,2009. IEEE,2009:55-62
[87] Li Y, Fujimoto R M, Hunter M P. Online Travel Time Prediction Based onBoosting [C].200912th International IEEE Conference on IntelligentTransportation Systems (ITSC2009), St. Louis, MO,2009. IEEE,2009:759-764
[88] Li H, Zeng M, Lu M, et al. Adaboosting-based Dynamic WeightedCombination of Software Reliability Growth Models[J]. Quality andReliability Engineering International,2012,28(1):67-84
[89] Hao X, Meng F, Zhou Y. An Improved Adaboost.R Algorithm and ItsApplication in Mining Safety Monitoring [C]. Third InternationalSymposium on Intelligent Information Technology Application,2009(IITA2009), Nanchang, China,2009. IEEE,2009:287-290
[90] Xu R Z, Geng X F, Zhou F Y. A Short Term Load Forecasting Based onBagging-ELM Algorithm [C]. Proceedings of the2012InternationalConference on Information Technology and Software Engineering, Beijing,2013. Springer Berlin Heidelberg:507-514
[91] Gao F, Kou P, Gao L, et al. Boosting Regression Methods Based on aGeometric Conversion Approach: Using SVMs Base Learners [J].Neurocomputing,2013,113:67-87
[92]赵学智,叶邦彦. SVD和小波变换的信号处理效果相似性及其机理分析[J].电子学报,2008,36(08):1582-1589
[93]张波,李健君.基于Hankel矩阵与奇异值分解(SVD)的滤波方法以及在飞机颤振试验数据预处理中的应用[J].振动与冲击,2009,28(02):162-166+208
[94] Jha S K, Yadava R D S. Denoising by Singular Value Decomposition and ItsApplication to Electronic Nose Data Processing [J]. Sensors Journal, IEEE,2011,11(1):35-44
[95] Zhao X Z, Ye B Y. Selection of Effective Singular Values Using DifferenceSpectrum and Its Application to Fault Diagnosis of Headstock [J].Mechanical Systems and Signal Processing,2011,25(5):1617-1631
[96] Huang N E, Shen Z, Long S R, et al. The Empirical Mode Decompositionand the Hilbert Spectrum for Nonlinear and Non-Stationary Time SeriesAnalysis [J]. Proceedings of the Royal Society A: Mathematical, Physicaland Engineering Sciences,1998,454(1971):903-995
[97] Flandrin P, Gonc alves P, Rilling G. Detrending and Denoising withEmpirical Mode Decompositions [C]. Proceedings of the12th EuropeanSignal Processing Conference (EUSIPCO'04), Vienna, Austria,2004.2004:1581-1584
[98] Wu Z, Huang N E, Long S R, et al. On the Trend, Detrending, and Variabilityof Nonlinear and Nonstationary Time Series [J]. Proc Natl Acad Sci U S A,2007,104(38):14889-14894
[99] Rilling G, Flandrin P, Gon alvés P. On Empirical Mode Decomposition andits Algorithms [C]. IEEE-EURASIP Workshop on Nonlinear Signal andImage Processing (NSIP’03), Grado, Italy,2003.2003:8-11
[100]丁刚,钟诗胜,栾圣罡.航空发动机滑油金属含量预测及其控制研究[J].润滑与密封,2006,(09):52-54
[101]欧阳楷，邹睿，刘卫芳.基于生物的神经网络的理论框架──神经元模型[J].北京生物医学工程,1997,(02):93-101
[102] Hagan M T, Menhaj M B. Training Feedforward Networks with theMarquardt Algorithm [J]. IEEE Transactions on Neural Networks,1994,5(6):989-993
[103] Mackey M C, Glass L. Oscillation and Chaos in Physiological ControlSystems [J]. Science,1977,197(4300):287-289
[104] Schapire R E. The Strength of Weak Learnability [J]. Machine learning,1990,5(2):197-227
[105] Polikar R. Ensemble Based Systems in Decision Making [J]. Circuits andSystems Magazine, IEEE,2006,6(3):21-45
[106] Solomatine D P, Shrestha D L. AdaBoost.RT: a Boosting Algorithm forRegression Problems [C].2004IEEE International Joint Conference onNeural Networks,2004. IEEE,2004:1163-1168
[107] Geman S, Bienenstock E, Doursat R. Neural Networks and the Bias/VarianceDilemma [J]. Neural computation,1992,4(1):1-58
[108] Ueda N, Nakano R. Generalization Error of Ensemble Estimators [C]. IEEEInternational Conference on Neural Networks,1996, Washington, DC,1996.IEEE,1996:90-95
[109] Zhou Z H. Ensemble Methods: Foundations and Algorithms [M]. Chapman&Hall,2012:99-102
[110] Solomatine D P. Local and Hybrid Learning Models in Forecasting NaturalPhenomena [C]. Proceedings of the International Joint Conference on NeuralNetworks (IJCNN),2005. IEEE,2005:1716-1721
[111] Yoav F, Schapire R E. Experiments with a New Boosting Algorithm [C].Machine Learning: Proceedings of the Thirteenth Conference,1996. MorganKaufmann Publishers, Inc,1996:148-156
[112] Drucker H. Improving Regressors using Boosting Techniques [C]. ICML '97Proceedings of the Fourteenth International Conference on MachineLearning, San Francisco, CA, USA,1997. San Francisco, CA, USA, MorganKaufmann Publishers Inc,1997:107-115
[113] Yang Y H, Lin Y C, Su Y F, et al. Music Emotion Classification: ARegression Approach [C].2007IEEE International Conference onMultimedia and Expo, Beijing, China,2007. IEEE,2007:208-211
[114] Tian H X, Mao Z Z. An Ensemble ELM Based on Modified AdaBoost.RTAlgorithm for Predicting the Temperature of Molten Steel in Ladle Furnace[J]. IEEE Transactions on Automation Science and Engineering,2010,7(1):73-80
[115] Costa E O, Pozo A, Vergilio S R. Using Boosting Techniques to ImproveSoftware Reliability Models Based on Genetic Programming [C].ICTAI-2006: Eighteenth International Conference on Tools with ArtificialIntelligence, Proceedings, Arlington, VA,2006. IEEE,2006:643-650
[116] Bar A, Rokach L, Shani G, et al. Improving Simple Collaborative FilteringModels Using Ensemble Methods [C].11th International Workshop, MCS2013, Nanjing, China,2013. Springer Berlin Heidelberg:1-12
[117] Shrestha D L, Solomatine D P. Experiments with AdaBoost.RT, an ImprovedBoosting Scheme for Regression [J]. Neural Comput,2006,18(7):1678-1710
[118] Moreiraa J M, Soaresb C, Jorgeb A M, et al. Ensemble Approaches forRegression: A Survey [J]. ACM Computing Surveys,2007,45(1):
[119] Yeung D S, Chan P P. A Novel Dynamic Fusion Method Using LocalizedGeneralization Error Model [C]. IEEE International Conference on Systems,Man and Cybernetics, San Antonio, TX,2009. IEEE,2009:623-628
[120] Cao D S, Xu Q S, Liang Y Z, et al. The Boosting: A New Idea of BuildingModels [J]. Chemometrics and Intelligent Laboratory Systems,2010,100(1):1-11
[121] Niall R, Patterson D. A Fusion of Stacking with Dynamic Integration [C].20th International Joint Conference on Artificial Intelligence, Hyderabad,India,2007. Morgan Kaufmann Publishers Inc,2007:2844-2849
[122] Puuronen S, Terziyan V, Tsymbal A. A Dynamic Integration Algorithm for anEnsemble of Classifiers [C].11th International Symposium, Foundations ofIntelligent Systems Warsaw, Poland,1999. Springer Berlin Heidelberg,1999:592-600
[123] Alexey T, Puuronen S. Bagging and Boosting with Dynamic Integration ofClassifiers [C].4th European Conference, Principles of Data Mining andKnowledge Discovery Lyon, France,2000. Springer Berlin Heidelberg,2000:116-125
[124] Jo o M M, Jorge A M, Soares C, et al. Ensemble Learning: A Study onDifferent Variants of the Dynamic Selection Approach [C].6th InternationalConference, Machine Learning and Data Mining in Pattern Recognition,Leipzig, Germany,2009. Springer Berlin Heidelberg,2009:191-205
[125] Niall R, Patterson D, Anand S, et al. Dynamic Integration of RegressionModels [C]. Proceedings of the5th International Multiple Classifier SystemsWorkshop, Cagliari, Italy,2004. Springer Berlin Heidelberg,2004:164-173
[126] Alexey T, Puuronen S, Terziyan V. A Technique for Advanced DynamicIntegration of Multiple Classifiers [C]. Proceedings of the FinnishConference on Artificial Intelligence,1998.1998:71-79
[127] Huang G B, Zhu Q Y, Siew C K. Extreme Learning Machine: Theory andApplications [J]. Neurocomputing,2006,70(1-3):489-501
[128] Fukunaga K, Narendra P M. A Branch and Bound Algorithm for ComputingK-Nearest Neighbors [J]. IEEE Transactions on Computers,1975,100(7):750-753
[129] Ko A H, Sabourin R, Britto Jr A S. From Dynamic Classifier Selection toDynamic Ensemble Selection [J]. Pattern Recognition,2008,41(5):1718-1731
[130] Woods K, Kegelmeyer W P, Bowyer K. Combination of Multiple ClassifiersUsing Local Accuracy Estimates [J]. IEEE Transactions on Pattern Analysisand Machine Intelligence,1997,9(4):405-410
[131] Yeung D S, Ng W W, Wang D, et al. Localized Generalization Error Modeland Its Application to Architecture Selection for Radial Basis FunctionNeural Network [J]. IEEE Trans Neural Netw,2007,18(5):1294-1305
[132] Zhang C X, Zhang J S. A Local Boosting Algorithm for SolvingClassification Problems [J]. Computational Statistics&Data Analysis,2008,52(4):1928-1941
[133] Huang G B, Wang D H, Lan Y. Extreme Learning Machines: A Survey [J].International Journal of Machine Learning and Cybernetics,2011,2(2):107-122
[134] Huang G B, Zhou H M, Ding X J, et al. Extreme Learning Machine forRegression And Multiclass Classification [J]. IEEE Transactions on Systems,Man, and Cybernetics, Part B: Cybernetics,2012,42(2):513-529
[135] http://www.stat.wisc.edu/~reinsel/bjr-data/
[136] Khosravi A, Nahavandi S, Creighton D, et al. Comprehensive Review ofNeural Network-Based Prediction Intervals and New Advances[J]. IEEETrans Neural Netw,2011,22(9):1341-1356
[137] Heskes T. Practical Confidence and Prediction Intervals [J]. Advances inneural information processing systems1997:176-182
[138] Khosravi A, Nahavandi S, Creighton D. Construction of Optimal PredictionIntervals for Load Forecasting Problems [J]. IEEE Transactions on PowerSystems,2010,25(3):1496-1503
[139] Zio E. A Study of the Bootstrap Method for Estimating the Accuracy ofArtificial Neural Networks in Predicting Nuclear Transient Processes [J].IEEE Transactions on Nuclear Science,2006,53(3):1460-1478
[140] Oleng N O, Gribok A, Reifman J. Error Bounds for Data-Driven Models ofDynamical Systems [J]. Comput Biol Med,2007,37(5):670-679
[141] De Veaux R D, Schumi J, Schweinsberg J, et al. Prediction Intervals forNeural Networks Via Nonlinear Regression[J]. Technometrics,1998,40(4):273-282
[142] MacKay D J. The Evidence Framework Applied to Classification Networks[J]. Neural computation1992,4(5):720-736
[143] Nix D A, Weigend A S. Estimating the Mean and Variance of the TargetProbability Distribution [C].1994. IEEE World Congress on ComputationalIntelligence,1994IEEE International Conference on Neural Networks,Orlando, FL,1994. IEEE,1994:55-60
[144] Khosravi A, Nahavandi S, Creighton D, et al. Optimizing The Quality ofBootstrap-based Prediction Intervals [C].2011International JointConference on Neural Networks (Ijcnn), San Jose, CA,2011. IEEE,2011:3072-3078
[145] Riedmiller M, Braun H. A Direct Adaptive Method for FasterBackpropagation Learning: the Rprop Algorithm [C]. IEEE InternationalConference on Neural Networks, San Francisco, CA,1993. IEEE,1993:586-591
[146] Ahmad F, Isa N A M, Osman M K, et al. Performance Comparison ofGradient Descent and Genetic Algorithm Based Artificial Neural NetworksTraining [C]. IEEE10th International Conference on Intelligent SystemsDesign and Applications (ISDA), Cairo,2010. IEEE,2010:604-609
[147] Anastasiadis A D, Magoulas G D, Vrahatis M N. New Globally ConvergentTraining Scheme Based on the Resilient Propagation Algorithm [J].Neurocomputing,2005,64:253-270
[148] Santosh T V, Vinod G, Saraf R K, et al. Application of Artificial NeuralNetworks to Nuclear Power Plant Transient Diagnosis[J]. ReliabilityEngineering&System Safety,2007,92(10):1468-1472
[149] Nada R H, Habashy D M, El-Salam F A, et al. Modeling of Aging Processfor Supersaturated Solution Treated of Al–3wt%Mg Alloy [J]. MaterialsScience and Engineering: A,2013,567(0):80-83
[150] Gebraeel N, Elwany A, Jing P. Residual Life Predictions in the Absence ofPrior Degradation Knowledge [J]. IEEE Transactions on Reliability,2009,58(1):106-117