Condition‐based Maintenance Optimization Using Neural Network‐based Health Condition Prediction

Artificial neural network (ANN)‐based methods have been extensively investigated for equipment health condition prediction. However, effective condition‐based maintenance (CBM) optimization methods utilizing ANN prediction information are currently not available due to two key challenges: (i) ANN prediction models typically only give a single remaining life prediction value, and it is hard to quantify the uncertainty associated with the predicted value; (ii) simulation methods are generally used for evaluating the cost of the CBM policies, while more accurate and efficient numerical methods are not available, which is critical for performing CBM optimization. In this paper, we propose a CBM optimization approach based on ANN remaining life prediction information, in which the above‐mentioned key challenges are addressed. The CBM policy is defined by a failure probability threshold value. The remaining life prediction uncertainty is estimated based on ANN lifetime prediction errors on the test set during the ANN training and testing processes. A numerical method is developed to evaluate the cost of the proposed CBM policy more accurately and efficiently. Optimization can be performed to find the optimal failure probability threshold value corresponding to the lowest maintenance cost. The effectiveness of the proposed CBM approach is demonstrated using two simulated degradation data sets and a real‐world condition monitoring data set collected from pump bearings. The proposed approach is also compared with benchmark maintenance policies and is found to outperform the benchmark policies. The proposed CBM approach can also be adapted to utilize information obtained using other prognostics methods. Copyright © 2012 John Wiley & Sons, Ltd.     

Condition based maintenance optimization for multi-component systems using proportional hazards model

The objective of condition based maintenance (CBM) is typically to determine an optimal maintenance policy to minimize the overall maintenance cost based on condition monitoring information. The existing work reported in the literature only focuses on determining the optimal CBM policy for a single unit. In this paper, we investigate CBM of multi-component systems, where economic dependency exists among different components subject to condition monitoring. The fixed preventive replacement cost, such as sending a maintenance team to the site, is incurred once a preventive replacement is performed on one component. As a result, it would be more economical to preventively replace multiple components at the same time. In this work, we propose a multi-component system CBM policy based on proportional hazards model (PHM). The cost evaluation of such a CBM policy becomes much more complex when we extend the PHM based CBM policy from a single unit to a multi-component system. A numerical algorithm is developed in this paper for the exact cost evaluation of the PHM based multi-component CBM policy. Examples using real-world condition monitoring data are provided to demonstrate the proposed methods.     

Decision Support System for Maintenance Management Using Bayesian Networks

The maintenance process has undergone several major developments that have led to proactive considerations and the transformation fiom the traditional ＂fail and fix＂ practice into the ＂predict and prevent＂ proactive maintenance methodology. The anticipation action, which characterizes this proactive maintenance strategy is mainly based on monitoring, diagnosis, prognosis and decision-making modules. Oil monitoring is a key component of a successful condition monitoring program. It can be used as a proactive tool to identify the wear modes of rubbing pans and diagnoses the faults in machinery. But diagnosis relying on oil analysis technology must deal with uncertain knowledge and fuzzy input data. Besides other methods, Bayesian Networks have been extensively applied to fault diagnosis with the advantages of uncertainty inference; however, in the area of oil monitoring, it is a new field. This paper presents an integrated Bayesian network based decision support for maintenance of diesel engines.     

Maintenance scheduling for a manufacturing system of machines with adjustable throughput

This paper considers the problem of analyzing and optimizing joint schedules of maintenance and throughput adjustment operations in manufacturing systems. The purpose of joint scheduling of maintenance and throughput changing operations is to maximize the cost benefits of maintenance operations in manufacturing systems in which some or all of the machines can execute their function under different process settings, resulting in different machine and system throughputs. Such a capability enables one to strategically slow down more degraded machines or accelerate freshly maintained machines so that production targets can be met and maintenance operations can be offset to times when they are less intrusive on the manufacturing process. A Monte-Carlo-simulation-based method is proposed for the evaluation of cost effectiveness of any schedule of maintenance and throughput changing operations, and a genetic-algorithm-based method is proposed to enable searching for schedules that would maximize the cost benefits of these operations. A matrix chromosome representation of the joint schedules of maintenance and throughput adjustment operations is introduced and several mechanisms of chromosome evolution and selection are proposed and analyzed in numerical simulations of such manufacturing systems. Results indicate a good ability for the newly proposed methods to achieve a tradeoff between cost benefits of production and losses due to maintenance operations through strategic allocation of maintenance and throughput changing actions.     

Optimal Scheduling of Track Maintenance on a Railway Network

To ensure the safety and continued operation of the railway network system, many maintenance and renewal activities are performed on the track every month. Unplanned maintenance activities are expensive and would cause low service quality. Therefore, the track condition should be monitored, and when it has degraded beyond some acceptable limit, it should be scheduled for maintenance before failure. An optimal timetable of the maintenance activities is needed to be scheduled, planning the monthly workload, to reduce the effect on the transportation service and to reduce the potential costs. Considering the uncertainties of the deterioration process, the safety of transportation service, the lifetime loss of the replaced track, the maintenance cost and the travel cost, this article advances an optimisation model for the maintenance scheduling of a regional railway network. An enhanced genetic algorithm approach is proposed to search for a solution producing maintenance schedule such that the overall cost is minimised in a finite planning horizon. A case study is given to demonstrate the application of the method. The case study results were derived by using an enhanced genetic algorithm method, which is specifically developed to deal with the characteristics of the railway maintenance problem. Copyright © 2012 John Wiley & Sons, Ltd.     

以可靠性为中心的维修方法及其在隧道中的应用

本文针对目前隧道监控维修状况及存在的问题,提出以可靠性为中心的维修(RCM)作为维修理念,并将其引入到隧道管理中,对隧道设备的故障模式及可靠性进行分析,并对RCM方法在隧道监控维修中的应用做了初步研究。最后通过案例说明,隧道采用RCM方法可有效降低维修费用,具有一定的经济效益和社会效益。     

Modeling and Analysis of Complex Equipment Maintenance Dynamics

An equipment maintenance system is naturally a complex dynamical system. The effective mamanagement must be based on the knowledge of the system＇s intrinsic dynamics. And the strueture of the maintenance system determines its behavior. This paper analyzes the basic structures and elements of a maintenance system for complex multi-components equipment. The maintenance system is considered as a dynamic system whose behavior is influenced by its structure＇s feedback and interaction, and the system＇s available resources. Building the dynamical model with Simulink, we show some results about the maintenance system＇s nonlinear dynamics, ods. The model can be used for understanding and which operational adjustments of maintenance which are never given by stochastic process methdetermining maintenance system behavior, towards n of maintenance requirements and timely supply of maintenance resources can be made in a more informed way.     

考虑预防维修时间及相关性的多部件系统维护策略

针对动车组关键系统维修过程中涉及部件数量多、维修时间长、维修费用高的特点，提出了预防维修时间、故障相关性、经济相关性3重因素影响下的多部件系统机会维护策略。首先对多部件建立部件故障率模型。在此基础上，考虑预防维修时间，将部件层预防维修成本分为独立时间成本和依赖时间成本，对经济相关性进行建模，再以系统维修费用率最小为优化目标建立系统层维护模型，并应用遗传算法求解。最后通过算例表明，相较于考虑单一因素的维修策略，所提维修策略可降低系统维修成本10%及以上，验证了考虑维修时间和部件联合相关性的必要性和有效性。     

维修工程管理研究与发展综述

综述了近几十年来维修工程管理的研究与发展，如人工智能、故障诊断、机器状态监测技术(如振动分析、红外检测、热录像仪等)、预测和预防性维修(PPM)、全员生产维修(TPM)和主动性维修(PM)等，论述了综合应用机器实时状态监测与故障诊断、人工智能、计算机通讯技术以及先进的维修管理理念的集成质量控制与维修系统，最后提出一种目前世界领先的远程智能维修系统。内容包括：实践应用中的维修管理评估、智能和集成维修管理、状态监测维修中的智能预测决策支持系统(IPDSS)、设备状况衰退趋势预测——人工神经网络方法、IPDSS支持的维修管理、故障诊断中的人工智能应用、基于可靠性的预防性维修安排和远程智能维修系统。     

Maintenance contract assessment for aging systems

This paper considers an aging system, where the system failure rate is known to be an increasing function. After any failure, maintenance is performed by an external repair team. Repair rate and cost of repair are determined by a corresponding maintenance contract with a repair team. There are many different maintenance contracts suggested by the service market to the system owner. In order to choose the best maintenance contract, a total expected cost during a specified time horizon should be evaluated for an aging system. In this paper, a method is suggested based on a piecewise constant approximation for the increasing failure rate function. Two different approximations are used. For both types of approximations, the general approach for building the Markov reward model is suggested in order to assess lower and upper bounds of the total expected cost. Failure and repair rates define the transition matrix of the corresponding Markov process. Operation cost, repair cost and penalty cost for system failures are taken into account by the corresponding reward matrix definition. A numerical example is presented in order to illustrate the approach. Copyright © 2008 John Wiley & Sons, Ltd.     

The Method of Mission Oriented Maintenance

In a real application,equipment in good condition sometimes have to be repaired because the equipment does not satisfy the requirements of the mission; However, it is not necessary to repair some equipment with failures because the failures do not affect the mission completion. In these cases, maintenance activity guided by present maintenance methods may sometimes affect the mission completion in time or bring about extra maintenance. To over-come the shortage of present maintenance methods, we propose an idea and method of mission oriented maintenance (MOM) to deal with the maintenance policy on these kinds of prob-lems. This method can work out different maintenance policies corresponding to different missions with full consideration of missions and mission requirements for the equipment.     

工程装备虚拟维修技术研究现状与展望

目的 针对工程装备设计的维修性难以验证,维修训练面临装备昂贵、不易重复拆卸、危险性高等问题,运用虚拟维修技术来评估装备的维修性和开展维修训练。方法 进行文献分析,理清虚拟维修的研究范畴和热点,从维修性设计、维修训练两个应用层面和虚拟人技术、虚拟样机及环境技术、维修过程与交互技术、维修性评估技术四个关键技术层面入手,对相关研究成果进行了整理和综述。结论 分析了目前研究存在的不足,提出了还可以进一步发展基于传感器技术的沉浸式交互方式,完善核心软件技术,研究多人协同维修与团体评估方法。     

不确定环境下复杂机械装配系统设备维修方法

为保障不确定环境下复杂机械装配系统的连续性并降低其维修成本,提出一种以马尔可夫决策理论为基础的设备维修策略动态选择方法。在综合考虑系统运行成本、缓冲库存成本、设备维修成本及停机损失成本的基础上,构建了装配系统可靠性成本模型。该模型以带有中间缓冲区的二级装配系统为研究对象,以设备状态和缓冲库存量为自变量,以可靠性成本为目标函数。分析了装配系统的不同运行状态,利用模拟退火算法和模糊非线性混合整数目标规划对可靠性成本模型求解,制定装配系统最优维修方法。该方法降低了装配系统停机时间,减少了设备维修次数,可为生产线设计和维修计划的制定提供依据。最后,通过算例分析验证了模型的有效性和可行性。     

The State of the Art of Hidden Markov Models for Predictive Maintenance of Diesel Engines

The maintenance of diesel Engines is usually scheduled according to the maintenance procedures defined by manufacturers. However, the state of the art shows that the condition monitoring maintenance associated with adequate prediction algorithms allows performance improvement both by increasing the intervals between interventions and by helping to maintain reliability levels. There are many types of variables that can be used to measure equipment condition, as is the case of several types of pollutant emissions such as NO_x, CO2, HC, PM, and NOISE, among others. This is a typical problem that can be solved through a hidden Markov model, taking into account the specificity of this type of equipment. The paper describes two algorithms that can help to increase the quality of assessment of engine states and the efficiency of maintenance planning. Those are the Viterbi and Baum–Welch algorithms. The importance of how to calculate the performance index of the model by the use of the perplexity algorithm is also emphasized. In this paper, a new paradigm is proposed, designated as ecological predictive maintenance. Copyright © 2017 John Wiley & Sons, Ltd.     

Maintenance schedule optimisation for a railway power supply system

High reliability is the crucial requirement in railway operation and a power supply system is one of the key components of electrified railways. The cost-effectiveness of the maintenance works is also the concern of the railway operators while the time window on trackside maintenance is often limited. Maintenance scheduling is thus essential to uphold reliability and to reduce operation cost. It is however difficult to formulate the optimal schedule to meet both reliability and maintenance cost for a railway power supply system as a whole because of its functional complexity and demanding operation conditions. Maintenance scheduling models to achieve reliability and maintenance cost are proposed in this study. Optimisation algorithms are then developed to attain the solutions of the model. The applicability of the models and efficiency of the solution algorithms are demonstrated in an example. The proposed methods are vitally important for the railway engineers and operators to assure the service quality in the increasing demands of the modern electrified railways.     

A study of time‐between‐events control chart for the monitoring of regularly maintained systems

Owing to usage, environment and aging, the condition of a system deteriorates over time. Regular maintenance is often conducted to restore its condition and to prevent failures from occurring. In this kind of a situation, the process is considered to be stable, thus statistical process control charts can be used to monitor the process. The monitoring can help in making a decision on whether further maintenance is worthwhile or whether the system has deteriorated to a state where regular maintenance is no longer effective. When modeling a deteriorating system, lifetime distributions with increasing failure rate are more appropriate. However, for a regularly maintained system, the failure time distribution can be approximated by the exponential distribution with an average failure rate that depends on the maintenance interval. In this paper, we adopt a modification for a time‐between‐events control chart, i.e. the exponential chart for monitoring the failure process of a maintained Weibull distributed system. We study the effect of changes on the scale parameter of the Weibull distribution while the shape parameter remains at the same level on the sensitivity of the exponential chart. This paper illustrates an approach of integrating maintenance decision with statistical process monitoring methods. Copyright © 2008 John Wiley & Sons, Ltd.     

Maintenance policies for systems with condition monitoring and obvious failures

We investigate a maintenance optimization problem with condition monitoring, which allows the decision maker to observe some wear-related variable throughout a system's lifetime to more accurately determine its degree of deterioration. Specifically, we examine the problem of adaptively scheduling observations (both perfect and imperfect) and preventive maintenance actions for a multistate, Markovian deterioration system with obvious failures, such that the long-run average-cost per unit time is minimized. We establish structural properties of the perfect observation-information problem and adjust them for heuristic use in the imperfect observation-information problem. We model both cases as partially observed Markov decision processes and provide numerical examples of optimal and heuristic solutions for both cases.     

Optimum Maintenance and Availability of Series Systems Subject to Imperfect Repair

ＯｐｔｉｍｕｍＭａｉｎｔｅｎａｎｃｅａｎｄＡｖａｉｌａｂｉｌｉｔｙｏｆＳｅｒｉｅｓＳｙｓｔｅｍｓＳｕｂｊｅｃｔｔｏＩｍｐｅｒｆｅｃｔＲｅｐａｉｒＨｏｎｇｚｈｏｕＷａｎｇＨｏａｎｇＰｈａｍＤｅｐａｒｔｍｅｎｔｏｆＩｎｄｕｓｔｒｉａｌＥｎｇｉｎｅｒｉｎｇ...     

Study on Equipment Selective Maintenance Strategy Based on Performance

1 IntroductionWith the development of science and technology, machine equipments become more and more complex. In China, engineers have found many disadvantages in the conventional time preventable maintenance of machines [1]. The essential thought of the strategy of conventional time preventable maintenance is to eliminate equipment's faults in earlier time through examining and repairing completely at fixed intervals. The main maintenance method is to disassemble an equipment and maintain i…     

A Production and Maintenance Planning Model with Restoration Cost Dependent on Detection Delay

In this paper, the joint problem of production planning and maintenance schedule is studied under the realistic assumption that the cost of process restoration is a function of the detection delay and the existence of shortages in the system. The detection delay is defined as the elapsed time since the production process has deteriorated until it is identified by some inspection procedure and repaired. Since production planning and maintenance problems have usually been studied as separate problems, this paper is an attempt to develop a formal framework for the joint problem. We have developed sufficient conditions for the optimality of the commonly used equal-interval maintenance schedule. The conditions are found to be a function of parameters such as the cost of defective items, the mean time for system deterioration, and the form of the restoration cost function. For specific restoration cost functions such as linear and exponential, an efficient solution procedure is presented for the simultaneous determination of the number of maintenance inspections in a production run, the length of the production run and consequently the economic manufacturing quantity, and the maximum level of backorders. A numerical example illustrates the use of this procedure and the differences between the optimal cost obtained by this procedure and the cost obtained by using the classical economic manufacturing quantity model.