机载液压系统寿命预测与估计

寿命预测和估计是机载液压系统健康管理的核心难点。综述了机载系统寿命预测与估计常用方法,针对长寿命机载系统不可能在出厂前给出准确寿命,提出动态数据更新的粒子滤波寿命估计方法。建立机载液压系统多场耦合作用下性能退化规律,将研制寿命试验的累积损伤表征在退化状态内部,实际飞行数据通过贝叶斯滤波动态更新到寿命估计模型中,考虑余度液压系统多退化状态增广,给出动态机载液压系统寿命预测与估计方法,实现机载液压系统高精度寿命估计。     

数字孪生驱动的航空发动机涡轮盘剩余寿命预测

为解决航空发动机涡轮盘剩余寿命在线预测难题,提出一种数字孪生驱动的涡轮盘剩余寿命预测方法。在建立数字孪生模型的过程中,首先,分析涡轮盘疲劳裂纹损伤机理,构建性能退化指标,建立涡轮盘性能退化过程的共性表征模型;其次,分析多种不确定性因素,采用状态空间模型建立涡轮盘性能退化过程的个性表征模型;然后,通过动态贝叶斯网络描述状态空间模型随时间的演化规律,建立涡轮盘性能退化过程的动态演化模型;最后,采用粒子滤波算法实现涡轮盘退化状态追踪和剩余寿命预测,从而完成涡轮盘性能退化数字孪生模型的建立。融合涡轮盘实时传感数据,通过贝叶斯推理实现对该数字孪生模型的动态更新。通过某型涡轮盘试验数据对该方法进行验证,结果表明该数字孪生模型能够较好地解决涡轮盘剩余寿命在线预测问题。     

动态状态空间模型及粒子滤波方法在滚动轴承寿命预测中的应用研究

针对滚动轴承物理模型难以准确建立,全寿命失效样本难以获取的问题,提出一种基于动态状态空间模型的滚动轴承寿命预测方法。该方法通过改进公式,构建模型参数定时更新的动态状态空间模型。将已知的滚动轴承运行状态数据输入动态状态空间模型,应用粒子滤波算法估计滚动轴承运行状态,实现滚动轴承寿命预测。运用滚动轴承全寿命实验数据对所提出方法进行验证,并将预测结果与Gamma模型预测结果对比分析,结果表明该方法优于Gamma模型预测方法,具有较强的工程实用性。     

动态贝叶斯网络在设备剩余寿命预测中的应用研究

采用动态贝叶斯网络对设备剩余寿命进行预测,建立了基于动态贝叶斯网络模型的设备剩余寿命预测框架模型,运用动态贝叶斯网络的粒子滤波近似推理算法对加工过程中钻头寿命预测进行实例研究,结果表明了该方法的有效性.     

锂离子电池循环寿命的融合预测方法

针对传统基于粒子滤波的锂离子电池剩余使用寿命预测方法的不足:过度依赖电池经验退化模型和模型输入变量单一的问题,提出了一种相关向量机、粒子滤波和自回归模型融合的锂离子电池剩余寿命预测的方法。通过相关向量机提取电池历史数据的退化趋势,构建趋势方程替换以往的电池经验退化模型,作为粒子滤波算法的状态转换方程。引入自回归模型的长期趋势预测值,替换观测值构建粒子滤波算法的观测方程。将3种方法相融合估计电池剩余寿命。实验结果表明:融合方法不仅预测精度高而且采用数据驱动的方法避免了构建复杂的电池机理退化模型,通用性强。     

基于数字孪生的光电探测系统性能预测

为解决光电探测系统由于多场耦合引起的性能退化建模难题,提出以光电探测系统的能量域表述为基础,建立数字孪生模型以解决其性能退化预测问题的方法。在该数字孪生模型的建立过程中,采用调制传递函数构建光电探测系统的静态性能物理模型;进一步采用动态贝叶斯网络表示调制传递函数随时间的演变规律,实现对其动态性能退化过程的图模型描述;最后通过粒子滤波算法实现系统的状态监测和性能预测,从而完成光电探测系统数字孪生模型的建立,并给出了具有不确定性估计的仿真验证结果,验证了数字孪生模型能够较好的解决光电探测系统的性能预测问题。     

基于扩展H∞粒子滤波算法的动力电池寿命预测方法

动力电池的性能随着使用会出现不可避免的老化,直接影响着电动汽车的性能和使用。在动力电池使用过程中对其进行剩余寿命的预测,可以确定动力电池的最佳维修和更换时机,进而有效延长动力电池寿命,增加电动汽车的续驶里程。因此,采用扩展H_∞粒子滤波算法进行动力电池的剩余寿命预测。进行锂离子动力电池循环老化试验,获取其全寿命周期的容量衰减数据。采用双指数拟合的方法建立电池容量衰减模型,并验证其准确性。将模型参数作为状态量,采用扩展H_∞粒子滤波算法对模型参数进行实时估计与更新,获得剩余循环次数以及预测结果的可信度。仿真结果表明,基于扩展H_∞粒子滤波算法得到的动力电池剩余寿命预测结果与基于粒子滤波得到的预测结果相比更加精确。     

基于随机滤波模型的齿轮箱剩余寿命预测研究

针对设备剩余寿命预测无法获取设备直接状态信息的问题,引入随机滤波模型,利用平时易于监测到的间接状态信息,来预测设备的剩余寿命.该模型采用贝叶斯递推理论,可以有效利用设备监测到的历史状态信息;针对小样本模型参数估计问题,采用主观数据和客观数据相结合的贝叶斯方法对模型的参数进行估计;最后,以齿轮箱全寿命实验为依据,利用该模...     

基于粒子滤波算法的动力电池SOC估计

针对电池荷电状态(SOC)容易受到电流、温度、循环寿命等非线性因素的影响,建立基于温度和电流变化的电池容量修正方程。结合安时法和复合电化学原理构建电池状态空间模型。由于粒子滤波算法对非高斯、非线性系统的适应性,因此选用粒子滤波算法来研究电池SOC估计。通过美国FTP-75工况和NEDC工况实验仿真显示,基于粒子滤波算法的电池SOC估计比扩展卡尔曼滤波算法估计精度高、适应性好。     

随机动态规划和粒子群嵌套寻优的PHEV能量优化

为了同时降低插电式混合动力汽车的能量消耗和电池寿命衰减速率,提出了随机动态规划和粒子群嵌套寻优的能量管理方法。建立了车辆的传动系统模型和电池模型,在电池寿命模型中引入寿命影响因子,使电池累计电量由理想情况转化为实际情况;为了描述车辆状态,建立了具有概率统计模型的驾驶循环模型;根据以上模型,将混合动力汽车能量控制问题转化为带约束优化问题;提出了随机动态规划和粒子群嵌套寻优的求解方法,使用粒子群搜索最优权重,达到能量消耗和寿命衰减速率最佳平衡。经仿真验证,相比于固定权重系数,嵌套寻优方法具有更优的控制结果;与文献[10]控制方法相比,等价燃油消耗减少了43.74%,电池寿命衰减率减少了35.53%,充分证明了嵌套寻优方的优越性。     

STOCHASTIC PROGNOSTICS FOR ROLLING ELEMENT BEARINGS

The capability to accurately predict the remaining life of a rolling element bearing is prerequisite to the optimal maintenance of rotating machinery performance in terms of cost and productivity. Due to the probabilistic nature of bearing integrity and operation condition, reliable estimation of a bearing's remaining life presents a challenging aspect in the area of maintenance optimisation and catastrophic failure avoidance. Previous study has developed an adaptive prognostic methodology to estimate the rate of bearing defect growth based on a deterministic defect-propagation model. However, deterministic models are inadequate in addressing the stochastic nature of defect-propagation. In this paper, a stochastic defect-propagation model is established by instituting a lognormal random variable in a deterministic defect-propagation rate model. The resulting stochastic model is calibrated on-line by a recursive least-squares (RLS) approach without the requirement of a priori knowledge on bearing characteristics. An augmented stochastic differential equation vector is developed with the consideration of model uncertainties, parameter estimation errors, and diagnostic model inaccuracies. It involves two ordinary differential equations for the first and second moments of its random variables. Solving the two equations gives the mean path of defect propagation and its dispersion at any instance. This approach is suitable for on-line monitoring, remaining life prediction, and decision making for optimal maintenance scheduling. The methodology has been verified by numerical simulations and the experimental testing of bearing fatigue life.     

一种基于空间卷积长短时记忆神经网络的轴承剩余寿命预测方法

传统的基于数据驱动的轴承剩余预测方法仍需要一定的先验知识,比如：特征指标选取、健康指标构建、失效阈值选定等等。预测结果严重依赖人工经验,为了克服这一缺点,基于深度学习方法提出了一种用于轴承剩余寿命预测的新方法,该方法的核心包括健康指标构建和剩余寿命计算。首先提出了一种无需先验知识的基于空间卷积长短时记忆神经网络（Convolutional long short-term memory neural network,ConvLSTM）的健康指标生成网络模型,该网络利用卷积神经网络的局部特征提取能力和长短时记忆网络的时间依赖特性,可直接从采集到的原始信号中挖掘反映退化程度的特征,构建健康指标,实现了高维原始数据向低维特征的映射转化,并利用Sigmoid函数将其归至[0,1]区间内,实现了阈值的统一;然后,利用粒子滤波更新双指数寿命模型,实现剩余寿命结果的输出。利用轴承全寿命试验对所提方法进行了验证,并与其他相关方法进行对比,结果表明本文方法所构建的健康指标具有更好的趋势性、单调性和鲁棒性,同时剩余寿命预测的准确率更高。     

Sequential Monte Carlo Method Toward Online RUL Assessment with Applications

Online assessment of remaining useful life(RUL) of a system or device has been widely studied for performance reliability, production safety, system conditional maintenance, and decision in remanufacturing engineering. However,there is no consistency framework to solve the RUL recursive estimation for the complex degenerate systems/device.In this paper, state space model(SSM) with Bayesian online estimation expounded from Markov chain Monte Carlo(MCMC) to Sequential Monte Carlo(SMC) algorithm is presented in order to derive the optimal Bayesian estimation.In the context of nonlinear non-Gaussian dynamic systems, SMC(also named particle filter, PF) is quite capable of performing filtering and RUL assessment recursively. The underlying deterioration of a system/device is seen as a stochastic process with continuous, nonreversible degrading. The state of the deterioration tendency is filtered and predicted with updating observations through the SMC procedure. The corresponding remaining useful life of the system/device is estimated based on the state degradation and a predefined threshold of the failure with two-sided criterion. The paper presents an application on a milling machine for cutter tool RUL assessment by applying the above proposed methodology. The example shows the promising results and the effectiveness of SSM and SMC online assessment of RUL.     

Analysis of a robust Kalman filter in loosely coupled GPS/INS navigation system

GPS/INS integrated system is very subject to uncertainties due to exogenous disturbances, device damage, and inaccurate sensor noise statistics. Conventional Kalman filer has no robustness to address system uncertainties which may corrupt filter performance and even cause filter divergence. Based on the INS error dynamic equation, a robust Kalman filter is analyzed and applied in loosely coupled GPS/INS integration system. The norm bounded robust Kalman filter, with recursive form by solving two Riccati equations, guarantees a estimation variance bound for all the admissible uncertainties, and can evolve into the conventional Kalman filter if no uncertainties are considered. This paper will analyze the suitable case for the robust Kalman filter in GPS/INS system, the filter characteristics including parameter setting, parameter meaning, and filter convergence condition are discussed simutaneously. The robust filter performance will be compared with conventional Kalman filter through simulation results.     

Parametric adaptive estimation and backstepping control of electro-hydraulic actuator with decayed memory filter

Some unknown parameter estimation of electro-hydraulic system (EHS) should be considered in hydraulic controller design due to many parameter uncertainties in practice. In this study, a parametric adaptive backstepping control method is proposed to improve the dynamic behavior of EHS under parametric uncertainties and unknown disturbance (i.e., hydraulic parameters and external load). The unknown parameters of EHS model are estimated by the parametric adaptive estimation law. Then the recursive backstepping controller is designed by Lyapunov technique to realize the displacement control of EHS. To avoid explosion of virtual control in traditional backstepping, a decayed memory filter is presented to re-estimate the virtual control and the dynamic external load. The effectiveness of the proposed controller has been demonstrated by comparison with the controller without adaptive and filter estimation. The comparative experimental results in critical working conditions indicate the proposed approach can achieve better dynamic performance on the motion control of Two-DOF robotic arm.     

Distributed and decentralized state estimation in gas networks as distributed parameter systems

In this paper, a framework for distributed and decentralized state estimation in high-pressure and long-distance gas transmission networks (GTNs) is proposed. The non-isothermal model of the plant including mass, momentum and energy balance equations are used to simulate the dynamic behavior. Due to several disadvantages of implementing a centralized Kalman filter for large-scale systems, the continuous/discrete form of extended Kalman filter for distributed and decentralized estimation (DDE) has been extended for these systems. Accordingly, the global model is decomposed into several subsystems, called local models. Some heuristic rules are suggested for system decomposition in gas pipeline networks. In the construction of local models, due to the existence of common states and interconnections among the subsystems, the assimilation and prediction steps of the Kalman filter are modified to take the overlapping and external states into account. However, dynamic Riccati equation for each subsystem is constructed based on the local model, which introduces a maximum error of 5% in the estimated standard deviation of the states in the benchmarks studied in this paper. The performance of the proposed methodology has been shown based on the comparison of its accuracy and computational demands against their counterparts in centralized Kalman filter for two viable benchmarks. In a real life network, it is shown that while the accuracy is not significantly decreased, the real-time factor of the state estimation is increased by a factor of 10.