基于PCA-UPF的风力机轴承剩余寿命预测方法

为解决风力机轴承退化指标提取困难与剩余寿命预测精度低的问题,提出一种基于主成分分析(PCA)和无迹粒子滤波(UPF)的预测方法。该方法主要包括退化指标提取和寿命预测2个步骤。在退化指标提取部分,通过PCA对轴承实时振动信号的多域原始特征集进行融合,得到能够反映轴承衰退趋势的退化指标。在剩余寿命预测部分,通过对轴承历史数据的拟合分析构建退化模型,再利用UPF算法对模型参数进行更新,实现对轴承退化状态的跟踪和预测。使用实际风力机轴承监测数据对所提方法进行验证,结果表明该方法相比于传统的粒子滤波PF方法,能有效降低粒子退化程度,从而显著提高轴承剩余寿命预测精度,为大型风电机组的健康管理和可靠性评估提供参考依据。     

基于退化轨迹和Wiener模型的光伏组件剩余寿命预测方法

针对光伏组件退化过程呈现的非单调、随机特性以及对组件剩余寿命自适应预测的需求,建立基于维纳(Wiener)过程的退化模型,在此基础上,提出一种结合退化轨迹自适应更新光伏组件剩余寿命的方法。首先,构建基于Wiener过程的光伏组件功率退化模型,刻画组件退化过程的非单调性以及组件退化过程的时间不确定性和个体差异性;然后,基于光伏组件的退化轨迹,联合贝叶斯更新和期望最大化(EM)算法对模型参数进行实时自适应更新,并在此基础上预测光伏组件的剩余寿命分布。最后,通过比较不同方法下光伏组件剩余寿命预测值的误差,验证所提方法的可行性与优越性。     

基于双输入深度卷积神经网络的风力机轴承RUL预测

针对轴承剩余寿命预测中常用健康指标泛化性不足的问题，提出一种基于双输入深度卷积神经网络的轴承剩余寿命预测模型。首先使用自适应最大相关峭度解卷积方法处理轴承信号并采用特征融合手段得到信号的时间序列特征；然后，将信号的时频图和时间序列特征同时作为模型的输入，通过已建立的双输入深度卷积神经网络模型来预测轴承健康因子；最后使用门控循环单元网络与健康因子相结合的方法来预测轴承的剩余使用寿命。在公开的西安交通大学公布的XJTU轴承数据集上对所提方法进行验证，并在风力机高速轴轴承历史监测数据上进行应用。试验结果表明：该方法不但显著提升了健康因子的泛化性能，还在预测精度方面有优异表现。     

基于VF-DW-DFN的锂离子电池剩余寿命预测北大核心CSCD

锂离子电池作为各类储能系统与设备的重要组成部分,准确预测锂离子电池的剩余使用寿命对于保障电池相关产业和设施的可靠性与安全性起着关键作用。针对锂离子电池剩余寿命预测中存在的非平稳、非线性特性导致单一数据驱动方法的预测精度低、泛化性能差等问题,提出了一种基于变分滤波、数据规整和深度融合网络的数据驱动融合(VF-DW-DFN)方法。首先,利用变分滤波法去除原始电池退化序列中的随机噪声干扰,得到相对平稳的退化特征数据。然后,采用最优嵌入法构造预测滑窗,实现特征数据规整,减少信息损失。其次,设计了一种新型深度融合网络对电池非线性退化数据进行建模,辨识电池数据中的退化模式,实现最终的锂离子电池剩余寿命预测。最后,在钴酸锂锂离子电池数据集上进行了剩余寿命预测实验,实验预测的平均均方根误差为1.41%,平均剩余寿命绝对误差小于2个循环周期。实验结果表明所提出的方法泛化性能好,预测精度高,误差小,能够对锂离子电池的退化过程进行有效建模和准确预测。     

钠离子电池健康状态预测

钠离子电池健康状态(SOH)预测对于电池优化管理有重要意义,但由于钠离子电池老化机理复杂,影响因素众多,精准SOH预测挑战巨大.为此,本研究从健康状态时序测量数据出发,提出了基于双指数模型的粒子滤波法(DEM-PF)和基于小波分析的高斯过程回归法(WA-GPR),以实现钠离子电池单步SOH和剩余可用寿命(RUL)预测.前者直接采用双指数函数构建时序SOH数据模型,并结合PF算法进行模型参数更新;后者采用小波分析实现时序SOH数据多尺度解耦,采用GPR构建各尺度数据模型并进行融合后实施预测.实验结果表明,相比DEM-PF方法,WA-GPR方法的单步SOH和RUL预测效果更好,单步SOH预测均方根误差为0.8％,RUL预测误差最小为3次循环,从而为钠离子电池管理提供有效支撑.     

基于PCA-GPR的锂离子电池剩余使用寿命预测

从充电过程中的电压-容量曲线中提取出一个与电池寿命高度相关健康因子（HI）。然后利用主成分分析（PCA）对影响电池寿命的多维因素进行分析和降维,结合高斯过程回归（GPR）机器学习方法提出一个基于PCA-GPR的锂离子电池剩余使用寿命预测模型。最后进行锂离子电池剩余使用寿命预测并与PCA-BP神经网络、PCA-支持向量机（SVM）模型进行比较。结果表明,利用该文提出的HI及预测模型可有效提高锂离子电池剩余使用寿命预测精度,其中通过贝叶斯优化器优化后的PCA-GPR模型的预测效果最佳。     

基于图形特征的双输入卷积神经网络风力机轴承剩余寿命预测

提出一种基于图形特征的风力机轴承剩余使用寿命（RUL）预测方法。首先,基于连续小波变换（CWT）对时域振动数据样本集进行预处理,得到用于预测的时频图形数据集。然后,采用双输入卷积神经网络（DICNN）从图形数据集中提取特征映射,用于构造高性能健康指数（DICNN-HI）来表征轴承各退化阶段的状态。最后,结合DICNN-HI,采用基于高斯过程回归（GPR）的分析方法进行RUL预测,并用PRONOSTIA滚动轴承数据集进行验证。结果表明,该方法具有较高的健康指数预测精度,能有效反映滚动轴承的劣化状态,有助于实现风力机轴承的RUL预测。同时,也可为其他旋转机械设备的剩余寿命预测提供重要的理论参考,具有一定的实用价值。     

基于PSO优化的BP神经网络在电动机绝缘剩余寿命预测中的应用

针对普通的电动机绝缘剩余寿命预测模型收敛速度慢、结果偏差大的缺陷,提出了一种基于粒子群算法(PSO)优化BP神经网络的电动机绝缘剩余寿命预测模型。首先,利用PSO算法全局随机最优解搜索的特性,对传统BP神经网络模型的权值和阈值进行优化设计。其次,为便于预测模型的运算处理,对采集的三相异步电动机的数据进行归一化处理。最后,结合经PSO算法优化的BP神经网络模型对三相异步电动机的绝缘剩余寿命进行试验预测。结果表明,基于PSO优化的BP神经网络比传统BP神经网络有更为精准的预测能力以及更快的收敛速度。     

水电机组结构寿命评价预测软件研究

开发了水轮机组结构寿命评价预测软件,并针对水轮机内部故障多发位置“轴承-轴瓦”结构进行研究。对该结构危险点位置应力时间序列进行分析,得出其所受应力分布的概率密度函数;应用应力-强度干涉模型对该结构在指定剩余寿命下的失效概率进行预测,预测结果对水轮机的维护具有指导意义,有助于提高水轮机组运行的稳定性。     

基于集成学习的涡扇发动机剩余寿命预测模型

为了解决涡扇发动机的监测数据维数高、时间跨度长、给预测发动机剩余使用寿命带来困难的问题。本文提出了一种基于集成神经网络模型的发动机寿命预测系统，采用集成学习中的Stacking方法对单一的学习器进行集成来预测涡扇发动机的剩余使用寿命(RUL)。模型在NASA公共数据集C-MAPSS(Commercial Modular Aero-Propulsion System Simulation)上进行了发动机寿命预测实验验证，并与常用的机器学习方法和单一神经网络进行了比较。实验结果表明：模型在多种评价方法上综合表现最佳，且在超前预测上表现良好。     

Remaining useful life prediction of lithium-ion battery based on extended Kalman particle filter

Scientific estimation and prediction of the state of health (SOH) of lithium-ion battery, especially the remaining useful life (RUL), has important significance to guarantee the battery safety and reliability in the full life cycle to avoid catastrophic accidents as much as possible. In order to accurately predict the RUL of the lithium-ion battery, this paper firstly analyzes the problems of the standard particle filter (PF). Then, a novel extended Kalman particle filter (EKPF) is proposed, in which the extended Kalman filter (EKF) is used as the sampling density function to optimize PF algorithm. The life cycle tests are designed and carried out to get accurate and reliable data for the RUL prediction. And, the aging properties of lithium-ion battery are analyzed in detail. The RUL prediction is done based on the established capacity degradation model and the proposed EKPF method. Results show that the RUL prediction error of the proposed method is less than 5%, which has higher precision compared with the standard PF method and can be used both offline and online.     

Remaining useful life prediction of PEMFC based on cycle reservoir with jump model

Accurate prognosis of limited durability is one of the key factors for the commercialization of proton exchange membrane fuel cell (PEMFC) on a large scale. Thanks to ignoring the structure of the PEMFC and simplifying the prognostic process, the data-driven prognostic approaches was the commonly used for predicting remaining useful life (RUL) at present. In this paper, the proposed cycle reservoir with jump (CRJ) model improves the ESN model, changes the connection mode of neurons in the reservoir and speeds up the linear fitting process. The experiment will verify the performance of CRJ model to predict stacks voltage under static current and quasi-dynamic current conditions. In addition, the reliability of the CRJ model is verified with different amount of data as the training and test sets. The experimental results demonstrate that the CRJ model can achieve better effect in the remaining useful life prognosis of fuel cells.     

A novel health indicator for PEMFC state of health estimation and remaining useful life prediction

Proton Exchange Membrane Fuel Cell (PEMFC) is a promising renewable energy, while still limited by the short life duration. To postpone the end of life, approaches of health management and prognostic (PHM) are applied into the cells. The stack voltage and impedance are often used as the health indicator (HI) for estimating state of health (SoH) and predicting remaining useful life (RUL). However, on one hand, on-line measurement of impedance is hardly realizable while downtime measure costs a lot. On the other hand, a single HI based on voltage or impedance is difficult to express the degradation of PEMFC precisely. To tackle this problem, this paper develops a fusion HI and a prognostic methodology for PEMFC SoH estimation and RUL prediction. Moreover, geodesic distance is employed to estimate SoH. Afterwards, a 2nd order Gaussian degradation model is built to complete the RUL prognostics based on unscented particle filter (UPF). In the experiment, both mahalanobis distance and geodesic distance are employed to estimate the SoH based on the presented HI. Besides, a rational model is applied to predict the RUL compared with the proposed Gaussian model. Finally, the results show the efficiency and effectiveness of the SoH estimation and RUL prediction approaches based on the proposed HI.     

Online particle‐contaminated lubrication oil condition monitoring and remaining useful life prediction for wind turbines

The widespread deployment of industrial wind projects will require a more proactive maintenance strategy in order to be more cost competitive. This paper describes an ongoing research project on developing online lubrication oil condition monitoring and degradation detection tools using commercially available online sensors. In particular, an investigation on particle contamination of lubrication oil is reported. Methods are presented for online lubrication oil condition monitoring and remaining useful life prediction using viscosity and dielectric constant sensors along with particle filtering technique. Physical models are derived in order to establish the mathematical relationship between lubrication oil degradation and particle contamination level. Laboratory experiments are performed to validate the accuracy of the developed models by comparing viscosity and dielectric constant sensor outputs of different particle concentration levels with those simulated by the lubricant deterioration physical models. A case study on lubrication oil degradation detection and remaining useful life prediction is provided. Discussions on the potential for extrapolating the presented methods to typical wind turbine gearbox oil and the practical implementation of particle filter‐based approach for online wind turbine gearbox oil remaining useful life prediction are also included. Copyright © 2014 John Wiley & Sons, Ltd.     

A novel degradation model of proton exchange membrane fuel cells for state of health estimation and prognostics

The major challenges for the commercialization of proton exchange membrane fuel cells (PEMFCs) are durability and costs. Prognostics and health management technology is helpful to extend the lifetime and reduce the maintenance costs of PEMFCs. However, the common degradation model, especially in the model-based method and the hybrid method, has the disadvantages of low generality and accuracy. A novel degradation model is proposed by introducing a polarization resistance in this paper to overcome the above disadvantages. Combining the novel degradation model and particle filter, a model-based method is proposed to estimate the state of health (SOH) and predict the future degradation trend (FDT) and the remaining useful life (RUL) of PEMFCs. Then, two actual degradation datasets of PEMFCs are used to validate the model, the RUL errors of the novel degradation model on these two datasets are 12.6% and 12.7%, respectively, while the errors of the common degradation model are 17.8% and 33.4%, respectively. The results prove that the proposed degradation model has higher generality and accuracy than the common model.     

铺设滤料对无砂混凝土板过流能力的影响

为分析铺设滤料对无砂混凝土板过流能力的影响,通过模型试验和理论分析相结合的方法,研究不同铺设粒径、不同铺设厚度下的无砂混凝土板过流能力。结果表明,无砂混凝土板的过流能力受到铺设滤料粒径和厚度的影响,粒径越小、铺设厚度越大,其过流能力越差;但不论粒径和厚度如何变化,随着工作水头的不断增大,铺设有不同粒径和不同厚度滤料的无砂混凝土板的过流能力均呈逐渐增大的趋势;同时还发现流体在滤料和无砂混凝土板所组成的孔隙介质中流动时为紊流状态。运用量纲分析法确定了无砂混凝土板的过流流量公式,可为无砂混凝土板的工程实际应用提供理论依据和指导。     

基于不敏卡尔曼粒子滤波的动态电力负荷在线建模

电力负荷的时变性对电力系统实时动态仿真分析具有较大影响。为了提高实时动态仿真分析的精度，基于不敏卡尔曼粒子滤波提出一种动态电力负荷在线建模方法。针对一种指数型动态负荷模型结构，利用不敏卡尔曼粒子滤波算法对其参数进行在线辨识。通过这种方式，可以根据实时采集的量测数据在线修正动态负荷模型的参数，从而追踪电力负荷的实时变化特性。分别利用动态仿真平台和实际电力系统的量测数据进行仿真分析，结果表明了所提方法具有较高的在线参数辨识精度，并能对实际电力负荷的实时变化特性进行准确的描述。     

Performance degradation prediction of proton exchange membrane fuel cell using a hybrid prognostic approach

The proton exchange membrane fuel cell has been widely used for industrial systems; however, its performance gradually degrades during use. Therefore, the study on the performance degradation prediction of fuel cells is helpful to extend its lifespan. In this paper, a novel hybrid approach using a combination of model-based adaptive Kalman filter and data-driven NARX neural network is proposed to predict the degradation of fuel cells. The overall degradation trend (i.e., irreversible degradation process) is captured by an empirical aging model and adaptive Kalman filter. Meanwhile, the detail degradation information (i.e., reversible degradation process) is depicted by the NARX neural network. Moreover, the correlation analysis of the reversible voltage time series is carried out to obtain the number of delays of the NARX neural network based on the autocorrelation function and the partial autocorrelation function. Then, the total degradation prediction is the sum of the overall degradation prediction and the detail degradation prediction. Finally, the prognostic capability of the proposed method is verified by two aging datasets, and the results show the effectiveness and superiority of the proposed method which can provide accurate degradation forecasting and remaining useful life.