融合多传感器数据的发动机剩余寿命预测方法

针对基于单一传感器数据的剩余寿命预测方法存在数据利用率低和预测精度不高的问题，论文提出了一种融合多传感器数据的发动机剩余寿命预测方法。首先将多个传感器数据融合成一个复合健康指标来表征发动机的退化性能，采用线性维纳过程对复合健康指标进行退化建模，通过极大似然估计方法确定模型参数，进而得到发动机的预测寿命。为了确定融合系数，提出了一种利用真实寿命与预测寿命的预测均方误差最小化的方法。融合系数确定后，基于训练发动机历史寿命数据，确定出模型参数的离线估计值；然后利用Bayesian公式，同时结合发动机的实时监测数据与参数的先验分布对模型参数进行实时更新，接着在首达时间的意义下推导出剩余寿命的概率分布，进而实现了发动机的剩余寿命在线预测。最后，选择商用模块化航空推进系统仿真数据集进行数值仿真实验，结果表明：相较于基于单一传感器的方法，论文所提方法能够提高剩余寿命预测的准确性，其剩余寿命预测的相对均方误差降低了2%左右。     

多源传感监测线性退化设备数模联动的剩余寿命预测方法

随着先进传感与监测技术的快速发展,实时获取随机退化设备的多源传感监测数据已成为现实,如何有效融合多源传感监测数据以实现随机退化设备剩余寿命的精准预测成为剩余寿命预测领域的研究前沿。针对多源传感监测的线性随机退化设备,提出了一种考虑随机失效阈值的数模联动剩余寿命预测新方法。该方法在离线训练过程中,基于多源传感历史数据提取的复合健康指标及据此线性随机退化建模预测的寿命,构建综合寿命预测值与设备实际寿命的均方误差及寿命预测方差的优化目标函数,形成复合健康指标提取与随机退化建模的反馈闭环,对多源传感器融合系数和复合健康指标对应的随机失效阈值分布参数进行优化调整,以实现复合健康指标提取与随机退化建模的自动匹配,即数模联动。在线预测时,根据提出的数模联动方法,融合实际运行设备的多源传感监测数据以获取复合健康指标,然后采用随机模型对其演变过程进行建模。同时,为使模型实时反映设备当前状况,提出了一种退化模型参数的贝叶斯更新方法,在此基础上基于首达时间得到了考虑设备失效阈值随机性的剩余寿命概率分布。最后,基于航空发动机的多源传感监测数据,验证了所提方法在改善复合健康指标特性和提高剩余寿命预测准确性方面的...     

基于KPCA-BLSTM的航空发动机多信息融合剩余寿命预测

复杂航空发动机在运行过程中易出现多退化信息而导致寿命预测不精确的问题,为此提出基于核主成分分分析（KPCA）和双向长短时记忆（BLSTM）神经网络的多信息融合寿命预测模型。首先采用KPCA 对多维退化数据集进行降维处理和信息融合,得到能够表征设备退化的低维特征数据集;然后利用BLSTM 神经网络模型对带有多维退化信息的航空发动机剩余寿命进行预测,得到监测数据与剩余寿命的映射关系;最后采用C-MAPSS 航空发动机退化数据集对提出的多信息融合寿命预测模型进行仿真验证,并与其他三种模型结果进行对比。结果表明：KPCA-BLSTM 能够对多维退化信息下的剩余寿命进行精准预测,本文提出的预测模型的误差与得分优于其他三种模型,而且预测精度更高。     

多重应力下滚动轴承剩余寿命预测

为了实现多重应力下滚动轴承的剩余寿命预测，有效利用不用应力下的退化数据，提出了一种基于加速模型和贝叶斯（Bayesian）理论的滚动轴承剩余寿命预测方法。通过拟合优度检验和威布尔（Weibull）概率图检验法对滚动轴承试验中的数据进行有效性分析。利用switching Kalman filters(SKF)判断滚动轴承各时刻的退化状态。当滚动轴承进入加速退化时，用指数模型拟合轴承退化过程，利用广义线性对数模型表示退化模型参数与应力的关系，根据修正后的轴承实时退化数据利用贝叶斯算法更新模型参数，得到滚动轴承剩余寿命的概率密度函数，从而实现滚动轴承剩余寿命预测。采用XJTU-SY轴承数据集进行验证，预测结果的均方根误差在20 min以内，证明该方法能够有效预测滚动轴承的剩余寿命。     

基于性能非线性退化统计模型的发动机剩余寿命预测

针对航空发动机在性能退化过程中普遍存在的非线性和不确定性问题,提出一种基于非线性退化数据的统计模型和剩 余寿命预测方法。通过对发动机性能真实退化轨道的分析,采用统计回归的建模方法建立发动机退化轨道模型,利用发动机的历 史数据,通过最小二乘估计求解模型中的未知参数;根据贝叶斯准则,以发动机实时监测数据与参数的先验分布对模型中的参数 进行实时更新,以发动机性能退化量首次达到红线值作为失效依据,采用蒙特卡洛仿真的方法得到发动机剩余寿命分布,实现了 对个体发动机剩余寿命的预测;通过试验数据进行发动机剩余寿命的预测,验证了该方法的准确性。结果表明：根据发动机退化 数据结合退化模型得到的个体发动机剩余寿命实时预测值末端均方根误差为0.02588,可以辅助指导维修决策。     

基于非线性Wiener过程航空发动机性能退化预测

针对线性随机过程航空发动机剩余使用寿命预测精度不高的问题,提出一种漂移系数为指数形式的非线性Wiener过程发动机性能退化建模,进而预测航空发动机的剩余寿命。基于直接监测发动机性能退化数据,构建发动机性能退化模型,根据Wiener过程首达阈值时间的数学性质,推导出剩余寿命的概率分布。通过极大似然估计构建退化模型中未知参数的似然函数,利用遗传算法得到发动机总体模型参数的离线估计值。考虑到不同发动机个体间的差异性,采用贝叶斯公式,结合发动机的实时监测数据与总体模型参数的先验分布对模型中随机参数进行实时更新,从而对个体发动机的剩余寿命实时预测。最后,选择商用航空发动机仿真数据集(C-MAPSS)进行实验,结果表明：针对个体发动机基于非线性随机过程方法,实时更新非线性Wiener方法能够提高航空发动机循环中期剩余寿命预测的准确性,提供更加可靠的预防性维修决策。     

基于退化失效与突发失效竞争的导弹剩余寿命预测

为了提高导弹剩余寿命预测结果的准确性,本文综合利用导弹的性能退化数据和突发失效时间数据,提出了基于退化失效和突发失效竞争的剩余寿命预测方法。在引入状态空间模型评估出整弹退化程度的基础上,采用Gamma过程建立退化失效模型;在假定突发失效概率与整弹性能退化程度相关的前提下,采用Weibull分布建立突发失效模型;进而建立退化失效与突发失效竞争模式下的导弹可靠度模型。案例应用证明了所提方法的有效性,对准确预测导弹剩余寿命,有效开展视情维修具有一定的工程价值。     

基于多头注意力机制的飞机发动机寿命预测研究

针对飞机发动机监测参数多和预测模型不能充分提取监测数据的有效信息等问题,基于一维卷积神经网络（1DCNN）、时序卷积神经网络（TCN）和多头注意力机制,提出一种新的网络结构以实现飞机发动机剩余寿命的准确预测。对多维特征参数分别建立一个1DCNN-TCN模型,利用两层1DCNN对飞机发动机的多元传感器信号进行特征提取,利用TCN对特征量的时序信息进行记忆,通过多头注意力机制对多个1DCNN-TCN的输出分别进行加权处理,并拼接最终结果。分析结果表明,采用本文方法得到的RMSE和Score值比目前文献中最优值分别降低了6.84%,63.41%。该方法显著提升了飞机发动机剩余寿命预测的准确性。     

基于半随机滤波-期望最大化算法的剩余寿命在线预测

剩余寿命(RL)预测是设备预测维护的关键环节。准确在线预测能够为维护策略的实时安排提供更加精确的技术支持,有效避免失效的发生。工程实际中,反映设备退化过程的性能指标往往不能直接监测,为解决隐含退化过程的剩余寿命在线预测问题,提出一种基于半随机滤波-期望最大化(EM)算法的预测方法。首先以剩余寿命为隐含状态,构建状态空间模型描述直接监测数据与设备剩余寿命间的随机关系。为实现单个设备剩余寿命的在线预测,依据到当前时刻为止的监测数据,采用扩展卡尔曼滤波(EKF)与期望最大化算法相互协作的方法实时估计与更新模型未知参数和剩余寿命分布。最后,将该方法用于惯性测量组合(IMU)剩余寿命在线预测问题,实验结果表明该方法能够提高预测的准确性并减少预测的不确定性。     

基于风扇退化的涡扇发动机寿命衰减双参数评估方法研究

由于涡扇发动机不同单元体之间存在耦合性,采用单一性能退化参数预测发动机剩余寿命明显是不全面的。本文根据风扇故障导致涡扇发动机退化机理,引入Frank Copula函数描述二元性能参数之间的相关性,并且采用二元非线性Wiener过程来构建性能退化模型,然后基于MCMC (Markov Chain Monte Carlo)方法进行模型参数估计,实现涡扇发动机剩余寿命预测。最终,通过涡扇发动机的仿真数据集来验证该方法的适用性。证明基于Copula函数的二元非线性Wiener过程建模为发动机剩余寿命预测提供了理论基础和技术支持。     

基于非线性Wiener过程的航空发动机性能衰减建模与剩余寿命预测

针对航空发动机在性能衰减过程中普遍存在的非线性和三源不确定性问题,提出了一种基于非线性Wiener过程的航空发动机性能衰减建模与剩余寿命（RUL）预测方法。首先,为解决目前大多数剩余寿命预测方法中潜在假设的局限性,即当前时刻估计的漂移系数与上一时刻漂移系数的后验估计完全相等,在状态空间模型的框架下建立了一类新的同时考虑非线性和三源不确定性的性能衰减模型,并在首达时间下推导出剩余寿命的分布。然后,针对新研发航空发动机缺乏历史数据和先验信息的问题,提出了一种基于Kalman滤波和条件期望最大化（ECM）算法的参数估计方法,使得估计的模型参数不依赖于历史数据量。同时能够在获得一个新的性能衰减数据后,实现对模型参数的自适应估计和在线更新,进而实时地更新航空发动机的剩余寿命分布。实验结果表明,本文方法可以有效地提高剩余寿命预测的准确性,能为航空发动机的维修决策提供可靠的依据。     

Remaining useful life prediction of aircraft lithium-ion batteries based on F-distribution particle filter and kernel smoothing algorithm

As an emergency and auxiliary power source for aircraft, lithium (Li)-ion batteries are important components of aerospace power systems. The Remaining Useful Life (RUL) prediction of Li-ion batteries is a key technology to ensure the reliable operation of aviation power systems. Particle Filter (PF) is an effective method to predict the RUL of Li-ion batteries because of its uncertainty representation and management ability. However, there are problems that particle weights cannot be updated in the prediction stage and particles degradation. To settle these issues, an innovative technique of F-distribution PF and Kernel Smoothing (FPFKS) algorithm is proposed. In the prediction stage, the weights of the particles are dynamically updated by the F kernel instead of being fixed all the time. Meanwhile, a first-order independent Markov capacity degradation model is established. Moreover, the kernel smoothing algorithm is integrated into PF, so that the variance of the parameters of capacity degradation model keeps invariant. Experiments based on NASA battery data sets show that FPFKS can be excellently applied to RUL prediction of Li-ion batteries.     

Remaining useful life prediction for engineering systems under dynamic operational conditions: A semi-Markov decision process-based approach

For critical engineering systems such as aircraft and aerospace vehicles, accurate Remaining Useful Life(RUL) prediction not only means cost saving, but more importantly, is of great significance in ensuring system reliability and preventing disaster. RUL is affected not only by a system's intrinsic deterioration, but also by the operational conditions under which the system is operating. This paper proposes an RUL prediction approach to estimate the mean RUL of a continuously degrading system under dynamic operational conditions and subjected to condition monitoring at short equi-distant intervals. The dynamic nature of the operational conditions is described by a discrete-time Markov chain, and their influences on the degradation signal are quantified by degradation rates and signal jumps in the degradation model. The uniqueness of our proposed approach is formulating the RUL prediction problem in a semi-Markov decision process framework, by which the system mean RUL can be obtained through the solution to a limited number of equations. To extend the use of our proposed approach in real applications, different failure standards according to different operational conditions are also considered. The application and effectiveness of this approach are illustrated by a turbofan engine dataset and a comparison with existing results for the same dataset.     

Aircraft air conditioning system health state estimation and prediction for predictive maintenance

The vast potential of system health monitoring and condition based maintenance on modern commercial aircraft is being realized through the innovative use of Airplane Condition Monitoring System (ACMS) data. However there are few methods addressing the issues of failure prognostics and predictive maintenance for commercial aircraft Air Conditioning System (ACS). This study developed a Bayesian failure prognostics approach using ACMS data for predictive maintenance of ACS. First, a health index characterizing the ACS health state is inferred from a multiple sensor signals using a data driven method. Then a dynamic linear model is proposed to describe the degradation process for failure prognostics. Bayesian inference formulas are carried out for degradation estimation and prediction. The developed approach is applied on a passenger aircraft fleet with ACMS data recorded for one year. The analysis of the case study shows that the developed method can produce satisfactory prognostics results, where all the ACS failure precursors are identified in advance, and the relative errors for the failure time prediction made when just entering the degradation warning stage are less than 8%. This would allow operators to proactively plan future maintenance.     

An adaptive-order particle filter for remaining useful life prediction of aviation piston pumps

An accurate estimation of the remaining useful life (RUL) not only contributes to an effective application of an aviation piston pump, but also meets the necessity of condition based maintenance (CBM). For the current RUL evaluation methods, a model-based method is inappropriate for the degradation process of an aviation piston pump due to difficulties of modeling, while a data-based method rarely presents high-accuracy prediction in a long period of time. In this work, an adaptive-order particle filter (AOPF) prognostic process is proposed aiming at improving long-term prediction accuracy of RUL by combining both kinds of methods. A dynamic model is initialized by a data-driven or empirical method. When a new observation comes, the prior state distribution is approximated by a current model. The order of the current model is updated adaptively by fusing the information of the observation. Monte Carlo simulation is employed for estimating the posterior probability density function of future states of the pump’s degradation. With updating the order number adaptively, the method presents a higher precision in contrast with those of traditional methods. In a case study, the proposed AOPF method is adopted to forecast the degradation status of an aviation piston pump with experimental return oil flow data, and the analytical results show the effectiveness of the proposed AOPF method.     

Remaining useful life prediction based on variation coefficient consistency test of a Wiener process

High-cost equipment is often reused after maintenance, and whether the information before the maintenance can be used for the Remaining Useful Life (RUL) prediction after the maintenance is directly determined by the consistency of the degradation pattern before and after the maintenance. Aiming at this problem, an RUL prediction method based on the consistency test of a Wiener process is proposed. Firstly, the parameters of the Wiener process estimated by Maximum Likelihood Estimation (MLE) are proved to be biased, and a modified unbiased estimation method is proposed and verified by derivation and simulations. Then, the h statistic is constructed according to the reciprocal of the variation coefficient of the Wiener process, and the sampling distribution is derived. Meanwhile, a universal method for the consistency test is proposed based on the sampling distribution theorem, which is verified by simulation data and classical crack degradation data. Finally, based on the consistency test of the degradation model, a weighted fusion RUL prediction method is presented for the fuel pump of an airplane, and the validity of the presented method is verified by accurate computation results of real data, which provides a theoretical and practical guidance for engineers to predict the RUL of equipment after maintenance.     

Remaining useful life prognostics for aeroengine based on superstatistics and information fusion

Remaining useful life（RUL） prognostics is a fundamental premise to perform conditionbased maintenance（CBM） for a system subject to performance degradation. Over the past decades,research has been conducted in RUL prognostics for aeroengine. However, most of the prognostics technologies and methods simply base on single parameter, making it hard to demonstrate the specific characteristics of its degradation. To solve such problems, this paper proposes a novel approach to predict RUL by means of superstatistics and information fusion. The performance degradation evolution of the engine is modeled by fusing multiple monitoring parameters, which manifest non-stationary characteristics while degrading. With the obtained degradation curve,prognostics model can be established by state-space method, and then RUL can be estimated when the time-varying parameters of the model are predicted and updated through Kalman filtering algorithm. By this method, the non-stationary degradation of each parameter is represented, and multiple monitoring parameters are incorporated, both contributing to the final prognostics. A case study shows that this approach enables satisfactory prediction evolution and achieves a markedly better prognosis of RUL.     

基于RF-SVR的燃油计量装置性能衰退检测和剩余寿命估计方法

为了实现航空发动机燃油系统的安全状态监测和健康管理,开展了燃油系统性能衰退检测和剩余使用寿命估计方面的研究。以燃油系统燃油计量装置为例,分析了其主要的性能衰退模式,设计了基于电流-速度数据的健康指标(HIs)选取方案,并考虑环境及模型参数不确定性,进行模型不确定性仿真,基于健康数据与性能衰退数据间的马氏距离对部件性能衰退进行检测。提出了基于随机森林-支持向量回归(RF-SVR)的剩余使用寿命(RUL)估计方法,利用通过RF特征选择优化的SVR模型实现部件RUL估计。最后基于某型民用涡扇发动机机械液压模型仿真数据对该方法进行了验证,结果表明:该方法的性能衰退检测虚警率及漏报率低于2%,RUL估计误差低于3%,可为航空发动机燃油系统的预测性维护提供参考。      

A new remaining useful life estimation method for equipment subjected to intervention of imperfect maintenance activities

As the key part of Prognostics and Health Management (PHM), Remaining Useful Life (RUL) estimation has been extensively investigated in recent years. Current RUL estimation studies considering the intervention of imperfect maintenance activities usually assumed that maintenance activities have a single influence on the degradation level or degradation rate, but not on both. Aimed at this problem, this paper proposes a new degradation modeling and RUL estimation method taking the influence of imperfect maintenance activities on both the degradation level and the degradation rate into account. Toward this end, a stochastic degradation model considering imperfect maintenance activities is firstly constructed based on the diffusion process. Then, the Probability Density Function (PDF) of the RUL is derived by the convolution operator under the concept of First Hitting Time (FHT). To implement the proposed RUL estimation method, the Maximum Likelihood Estimation (MLE) is utilized to estimate the degradation related parameters based on the Condition Monitoring (CM) data, while the Bayesian method is utilized to estimate the maintenance related parameters based on the maintenance data. Finally, a numerical example and a practical case study are provided to demonstrate the superiority of the proposed method. The experimental results show that the proposed method could greatly improve the RUL estimation accuracy for the degrading equipment subjected to imperfect maintenance activities.     

A data-driven health indicator extraction method for aircraft air conditioning system health monitoring

Prognostics and Health Management (PHM) has become a very important tool in modern commercial aircraft. Considering limited built-in sensing devices on the legacy aircraft model, one of the challenges for airborne system health monitoring is to find an appropriate health indicator that is highly related to the actual degradation state of the system. This paper proposed a novel health indicator extraction method based on the available sensor parameters for the health monitoring of Air Conditioning System (ACS) of a legacy commercial aircraft model. Firstly, a specific Airplane Condition Monitoring System (ACMS) report for ACS health monitoring is defined. Then a non-parametric modeling technique is adopted to calculate the health indicator based on the raw ACMS report data. The proposed method is validated on a single-aisle commercial aircraft widely used for short and medium-haul routes, using more than 6000 ACMS reports collected from a fleet of aircraft during one year. The case study result shows that the proposed health indicator can effectively characterize the degradation state of the ACS, which can provide valuable information for proactive maintenance plan in advance.