加速冲击损伤退化系统剩余寿命预测及预测维修决策

随着检测传感技术的发展,诸如风力发电机叶片等可对其状态进行检测,并依据检测结果进行剩余寿命预测.但此类系统在运行中受环境冲击影响较大,如何对冲击影响下的系统剩余寿命进行预测,并结合预测结果进行经济可靠的维修决策是一个值得研究的问题.对此,针对状态可检测的连续退化系统,研究考虑加速冲击损伤特性下的系统剩余寿命预测及基于预测的维修决策.首先,考虑自然退化和与退化相关的冲击损伤,构建加速冲击损伤退化模型和剩余寿命预测模型;其次,制定基于周期检测的状态维修与预测维修相结合的混合维修策略,并推导不同维修活动的发生概率;然后,构建以长期平均费用率最小为目标,以检测间隔和故障率阈值为决策变量的决策模型,并给出了优化解法;最后,以风力发电机叶片为案例验证模型的适用性和有效性,对系统的参数进行灵敏度分析,并与未考虑加速冲击损伤和未考虑预测的维修决策结果进行对比分析.     

Optimisation of PM scheduling for multi-component systems – a simulated annealing approach

Proper planning of preventive maintenance (PM) is crucial in many industries such as oil transmission pipelines, automotive and food industries. A critical decision in the PM plans is to determine frequencies and types of maintenance actions in order to achieve a certain level of system availability with a minimum total cost. In this paper, we consider the problem of obtaining availability-based non-periodic optimal PM planning for systems with deteriorating components. The objective is to sustain a certain level of availability with the minimal total maintenance-related costs. In the proposed approach, the planning horizon is divided into some inspection periods of equal intervals. For any given interval, a decision must be made to perform one of the three actions on each component; inspection, preventive repair and preventive replacement. Any of these activities has different effects on the reliability of the components and the corresponding distinct costs based on the required recourses. The cost function includes the cost for repair, replacement, system downtime and random failures. System availability and PM resources are the main constraints considered. Since the proposed model is combinatorial in nature involving non-linear decision variables, a simulated annealing algorithm is employed to provide good solutions within a reasonable time.     

The scheduling of maintenance. A resource-constraints mixed integer linear programming model

The scheduling of preventive maintenance is crucial in reliability and maintenance engineering. Hundreds of parts compose complex machines that require replacement and/or repairing. Maintenance involves the machine vendor (1), the machine user (2) and the service maintenance provider (3). The vendor and the maintenance service provider have to guarantee a high level of availability and productivity of the machines and maintain their down-time at a minimum even though they are installed worldwide and usually far from the vendor’s headquarters and/or the locations of the provider’s regional service offices. Moreover, many companies have great profits from maintenance and spare parts management.This study aims to illustrate an original mixed integer linear programming (MILP) model for the cost-based, reliability-based and resource-constraints scheduling of preventive maintenance actions. The model minimizes the total cost function made of spare parts contributions, the cost of the execution of the preventive actions and the cost of the additional repair activity in case of unplanned failure. The cost of the personnel of the producer and/or the maintenance service provider is also included. Finally, the paper presents a case study in a what-if environment demonstrating the effectiveness and the novelty of this study in real and complex applications.     

不完美维护下基于剩余寿命预测信息的设备维护决策模型

基于剩余寿命预测信息进行设备维护决策的研究中,现有方法通常仅考虑不完美维护对退化量或退化率的单一影响,忽略了不完美维护对两者的双重影响.鉴于此,针对随机退化设备,提出一种考虑不完美维护影响的性能退化模型与维护决策模型,融合了维护活动对设备退化量和退化率的双重影响.首先基于Wiener过程分阶段构建存在不完美维护干预的随机退化模型,在首达时间的意义下推导出剩余寿命的解析概率分布;然后基于剩余寿命的预测结果,以检测间隔和预防性维护阈值为决策变量建立维护决策模型;最后数值仿真实验验证了本文模型的有效性,并对费用参数进行了敏感性分析.实验结果表明本文模型具有潜在的工程应用价值.     

基于工程应用的备件分析及仿真研究

为解决型号研制中备件分析流程指导性不强、备件预测计算模型选控对比原则缺失、备件配置大量冗余的现状,从装备使用维护任务出发,开展基于工程应用的备件分析方法及仿真研究.依据行业标准规范,结合型号备件配置经验,制定基于工程应用的备件分析流程.以最少的保障资源需求满足装备固有的可靠性和安全性水平为前提,开展预防性维修备件分析,建立航空装备状态与保障资源的映射.选取适用于工程应用的修复性维修备件预测模型,基于Matlab仿真完成各预测模型选控对比研究.通过开展基于工程应用的备件分析及仿真研究,可实现航空装备备件的定性分析和定量计算,减少备件冗余,提高装备可用性和备件资源利用率.     

基于PHM的舰艇装备视情维修技术研究

针对PHM技术在舰艇装备视情维修中的应用需求,在分析剩余寿命预测机理的基础上,为解决舰艇装备历史数据少、状态变化普遍表现为非线性等难题,建立基于粒子滤波的多步剩余寿命预测模型,通过仿真表明：随着时间的推移剩余寿命PDF图变狭窄,剩余寿命预测误差逐渐降低,证明了预测方法的有效性。在此基础上,考虑装备维修费用最低为目标,建立以费用最低为目标的预防性维修时间确定模型,使得单位时间内的平均维修费用最低,并通过采用建立的基于粒子滤波的多步剩余寿命预测模型,最终确定了最优预防性维修时间,为装备保障人员拟订精确维修保障计划提供依据,也为舰艇装备开展视情维修奠定基础。     

Inspection and maintenance planning: an application of semi-Markov decision processes

In this paper, we consider a system which deteriorates stochastically from one time unit to another. The state of the system, which is a continuous random variable, can be observed only by inspection. The goal of this paper is to derive a predictive maintenance policy which indicates, at each inspection and according to the observed value, whether a preventive maintenance is necessary and when the next inspection should be done. The objective is to minimize the long-run average cost incurred by inspections, preventive maintenance and unexpected breakdowns. The problem is modelled by a semi-Markov decision process. Analytical properties are derived, and based on these properties, numerical methods are constructed to compute an optimal policy.     

成组更换策略下的备件库存决策仿真研究

针对成组更换策略下备件需求的特点,建立了无限使用期的成组更换策略与备件库存控制联合优化仿真模型。该模型通过模拟系统在预防维修间隔期T内的维修活动,得到1个预防维修间隔期内的总费用,然后在系统运行M(M足够大)个周期下,以期望单位时间总费用最小为目标,优化预防维修间隔期T和最大库存水平S。最后,基于案例,运用MATLAB对模型进行仿真优化计算,并分析了相关参数对模型优化结果的影响。     

A generalized sequential preventive maintenance policy for repairable systems with general random minimal repair costs

A generalized sequential preventive maintenance (PM) policy for repairable systems with general random minimal repair costs is proposed and analysed. After each (planned or unplanned) preventive maintenance, the system has a different failure distribution and the failure rate function increases with the number of preventive maintenances carried out. The criterion for the optimal policy is to minimize the expected cost per unit time for an infinite time span. It is shown that under certain reasonable assumptions, sequential preventive maintenance policy has unique solutions. Various special cases are considered.     

Discrete-time spare ordering policy with randomized lead times and discounting

The paper considers a generalized discrete‐time order‐replacement model for a single unit system, which is subject to random failure when in operation. Two types of discrete randomized lead times are considered for a spare unit; one is for regular (preventive) order and another is for expedited (emergency) order. The model is formulated based on the discounted cost criterion. The underlying two‐dimensional optimization problem is reduced to a simple one‐dimensional one and then the optimal ordering policy for the spare unit is characterized under two extreme conditions: (i) unlimited inventory time and (ii) zero inventory time for the spare unit. A numerical example is used to determine the optimal spare‐ordering policy numerically and to examine the sensitivity of the model parameters.     

Simultaneous optimization of repair and control-limit policy in condition-based maintenance

In condition-based maintenance (CBM) planning, collected information from system condition monitoring is the basis of making decision about conducting the maintenance and repair activities. Recently, ample number of studies has been conducted in CBM field especially, in control-limit policy. In control-limit policy, using proportional Hazards model and results of monitoring system condition, one can estimate hazard rate function and its condition’s transition probability matrix. Then, considering replacement costs, optimal control-limit can be determined minimizing the average cost in the long run. The presented model considers repair policy and their implementation cost, and the assumptions of repair during interval inspection is ignored. Then, a model is presented to determine the optimal control-limit and the best repair policy, in which the average total cost per unit time in the long-run, is minimized. At the end, a numerical example is illustrated.     

Discrete Event Logistics Systems (DELS) simulation modeling incorporating two-step Remaining Useful Life (RUL) estimation

Prognostics and Health Management (PHM) exerts an essential influence on the spare supply process and the maintenance activities. Discrete Event Logistics Systems (DELS) simulation model facilitates a better understanding of the maintenance and logistics/support systems. Previous DELS models treat the RUL estimation as a one shot event. However, the treatment would be rough to coordinate the logistics and maintenance activities, and the estimated RUL result would not be sufficiently reliable. In this paper, we propose the principle and operational technique of two-step RUL estimation for the DELS simulation model. Two-step RUL estimation starts with the component RUL modeling subject to a continuous accumulation of degradation. The component deterioration is modeled using a time-dependent stochastic process, which combines the linear degradation path with a random effect. Besides, the sequential logics of the DELS simulation model incorporating two-step RUL estimation is exploited in the local behavior study. Finally, the proposed technique is testified with a case study via the DELS simulation implementation, showing that the performance using two-step RUL estimation outperforms traditional one-step RUL estimation.     

A novel deep learning scheme for multi-condition remaining useful life prediction of rolling element bearings

The remaining useful life (RUL) prediction of a rolling element bearing is important for more reasonable maintenance of machinery and equipment. Generally, the information of a failure can hardly be acquired in advance while running and the degradation process varies in terms of different faults. Thus, fault identification is indispensable for a multi-condition RUL prediction, where, however, the fault identification and RUL prediction are separated in most studies. A new hybrid scheme is proposed in this paper for the multi-condition RUL prediction of rolling element bearings. The proposed scheme contains both classification and regression, where the 2D-DCNN based classifier and predictors are built concerning typical fault conditions of a bearing. For the online prediction, the raw signals are spanned in the time-frequency domain and then transferred into images as the input of the scheme. The classifier is used to monitor the vibration of rolling bearings for online fault recognition and excite the corresponding predictor for RUL prediction once a fault is detected. The output from the predictor is amended by the proposed adaptive delay correction method as the final prediction results. A demonstration is performed based on the XJTU-SY datasets and the results are compared with those from the state-of-the-art methods, which proves the superiority of the proposed scheme in improving the accuracy and linearity of RUL prediction. The time cost of the proposed online prediction scheme is also investigated and the results indicate high time effectiveness.     

Analysis of maintenance policies for finite life-cycle multi-state systems

Maintenance policies for multi-state systems (MSS) are often analyzed under infinite horizon assumptions. In practice, it is important to consider maintenance policies under a finite horizon because the life cycles of most systems are finite. In this paper, we consider a finite life-cycle MSS that is subject to both degradation and Poisson failures. We study two classes of maintenance policies – preventive replacements and corrective replacements, and their effectiveness in controlling the customer’s expected discounted maintenance cost (EDMC). For both policies, replacement decisions are modelled via two control parameters – a threshold on the current system state and a threshold on the residual life cycle, which is measured as the time span from present to the end of life cycle. We derive close-to-explicit forms of the cost models under each of the policy. Methodologies for optimizing the maintenance thresholds are further proposed. Computational results verify that preventive replacements outperform corrective replacements typically when the downtime cost per failure is relatively high compared to the repair cost.     

A mathematical model on EPQ for stochastic demand in an imperfect production system

In the paper, we develop an EPQ (economic production quantity) inventory model to determine the optimal buffer inventory for stochastic demand in the market during preventive maintenance or repair of a manufacturing facility with an EPQ (economic production quantity) model in an imperfect production system. Preventive maintenance, an essential element of the just-in-time structure, may cause shortage which is reduced by buffer inventory. The products are sold with the free minimal repair warranty (FRW) policy. The production system may undergo “out-of-control” state from “in-control” state, after a certain time that follows a probability density function. The defective (non-conforming) items in “in-control” or “out-of-control” state are reworked at a cost just after the regular production time. Finally, an expected cost function regarding the inventory cost, unit production cost, preventive maintenance cost and shortage cost is minimized analytically. We develop another case where the buffer inventory as well as the production rate are decision variables and the expected unit cost considering the above cost functions is optimized also. The numerical examples are provided to illustrate the behaviour and application of the model. Sensitivity analysis of the model with respect to key parameters of the system is carried out.     

基于混合果蝇优化算法的选址-库存联合优化策略

针对装备维修保障仓库系统运营费用高、仓库点位布局不合理、备件库存结构不合理等问题,建立以多品种联合补货问题为基础的装备维修备件仓库选址-库存控制决策联合优化模型,模型可用于求解仓库的开设位置、维修活动需求点的指派情况、仓库补货时间以及库存水平等.根据模型的结构特点,利用多种群协同进化的方法改进传统果蝇优化算法的位置更新方式,设计一种内外两层搜索策略的混合果蝇优化算法,外层搜索策略作为算法的主程序用于搜索仓库选址决策变量,内层搜索策略采用改进的RAND算法用于搜索库存控制决策变量.仿真结果表明,混合果蝇优化算法具有良好的求解效率,能够确保库存系统在一定服务水平的基础上有效降低库存运营总成本.     

Markov decision models for the optimal maintenance of a production unit with an upstream buffer

We consider a manufacturing system in which a buffer has been placed between the input generator and the production unit. The input generator supplies at a constant rate the buffer with the raw material, which is pulled by the production unit. The pull-rate is greater than the input rate when the buffer is not empty. The two rates become equal as soon as the buffer is evacuated. The production unit deteriorates stochastically over time and the problem of its optimal preventive maintenance is considered. Under a suitable cost structure it is proved that the optimal average-cost policy for fixed buffer size is of control-limit type, if the repair times are geometrically distributed. Efficient Markov decision process solution algorithms that operate on the class of control-limit policies are developed, when the repair times are geometrical or follow a continuous distribution. The optimality of a control-limit policy is also proved when the production unit after the end of a maintenance remains idle until the buffer is filled up. Furthermore, numerical results are given for the optimal policy if it is permissible to leave the production unit idle whenever it is in operative condition.     

A Risk-Averse Remaining Useful Life Estimation for Predictive Maintenance

Remaining useful life (RUL) prediction is an advanced technique for system maintenance scheduling. Most of existing RUL prediction methods are only interested in the precision of RUL estimation; the adverse impact of over-estimated RUL on maintenance scheduling is not of concern. In this work, an RUL estimation method with risk-averse adaptation is developed which can reduce the over-estimation rate while maintaining a reasonable under-estimation level. The proposed method includes a module of degradation feature selection to obtain crucial features which reflect system degradation trends. Then, the latent structure between the degradation features and the RUL labels is modeled by a support vector regression (SVR) model and a long short-term memory (LSTM) network, respectively. To enhance the prediction robustness and increase its marginal utility, the SVR model and the LSTM model are integrated to generate a hybrid model via three connection parameters. By designing a cost function with penalty mechanism, the three parameters are determined using a modified grey wolf optimization algorithm. In addition, a cost metric is proposed to measure the benefit of such a risk-averse predictive maintenance method. Verification is done using an aero-engine data set from NASA. The results show the feasibility and effectiveness of the proposed RUL estimation method and the predictive maintenance strategy.      

基于二阶段时间延迟的多产品生产系统生产与维修联合优化研究

针对多产品生产部件串联系统的生产和维修问题进行了研究,提出了基于二阶段时间延迟的联合优化模型。首先,基于生产周期分段理论,将整个周期等分成若干单位时间段,生产与维修共用每段时间,且若干时间段后采取一次预防维修。其次,考虑生产系统的实际生产时间、可用生产时间和维修耗费时间,建立了生产计划与维修计划总成本模型。其中,维修计划考虑缺陷和故障维修费用、维修检查费用,以及非正常状态下设备运行可能产生的不合格产品损失费用;生产计划考虑生产成本、库存成本、延期未交货成本和维修停机后恢复生产的设备启动成本。最后,通过算例分析,计算最优预防维修周期和各单位时间段各产品产量,验证了模型的有效性。     

基于故障率修正参数的生产设备预防性维修研究

点检定修制是生产设备预防性维修的1种重要方法,是全员、全过程对设备进行动态管理的1种设备管理方法。点检定修制的核心是确定点检周期,即预防性维修周期。为此引入故障率修正参数,建立了基于单位时间净生产效益的生产设备预防性维修模型,并通过仿真的方法确立最佳预防性维修周期和维修次数,阐述了维修程度的不同对预防性维修周期的影响以及预防性维修对企业增加利润的重要性。结果表明,选择合理的故障率修正参数可以有效降低故障率,增大平均单位产值,从而为管理者制定有效的维修计划提供决策。