Optimal maintenance policy for multi-component systems under Markovian environment changes

In this paper, we study multi-component systems, which environmental conditions and opportunistic maintenance (OM) involve. Environmental conditions will exert an influence on deterioration processes of the components in the system. For each component, the worse the environmental conditions are, the faster its deterioration speed is. We want to determine when to preventively maintain each component under such environmental influence. Our purpose is to minimize its long-run average maintenance cost. We decompose such a multi-component system into mutually influential single-component systems, and formulate the maintenance problem of each component as a Markov decision process (MDP). Under some reasonable assumptions, we prove the existence of the optimal (n_r, N_r) type policy for each component. A policy iteration method is used to calculate its optimal maintenance policy. Based on the method, we develop an iterative approximation algorithm to obtain an acceptable maintenance policy for a multi-component system. Numerical examples find that environmental conditions and OM pose significant effects on a maintenance policy.     

Optimal imperfect preventive maintenance policies for a used system

In some practical situations, it may be more economical to work a used system than do a new one. From this viewpoint, this article considers three basic preventive maintenance (PM) policies for a used system: the system with initial variable damage Y ₀ begins to operate at time 0, and suffers damage due to shocks. It fails when the total damage exceeds a failure level K and corrective maintenance is made immediately. To prevent such failure, it undergoes PM at a planned time T, a shock number N and a damage level k, but maintenances are imperfect. However, failure rate of a used system maybe higher than that of a new one, so some maintenance is applied to the policies at each shock in the extended models. Using the theory of cumulative processes, expected cost rate models are obtained, optimal policies which minimise them are derived analytically and discussed numerically.     

Optimum preventive maintenance policies for systems subject to random working times,replacement, and minimal repair

This paper proposes, from the economical viewpoint of preventive maintenance in reliability theory, several preventive maintenance policies for an operating system that works for jobs at random times and is imperfectly maintained upon failure. As a failure occurs, the system suffers one of two types of failure based on a specific random mechanism: type-I (repairable) failure is rectified by a minimal repair, and type-II (non-repairable) failure is removed by a corrective replacement. First, a modified random and age replacement policy is considered in which the system is replaced at a planned time T, at a random working time, or at the first type-II failure, whichever occurs first. Next, as one extended model, the system may work continuously for N jobs with random working times. Finally, as another extended model, we might consider replacing an operating system at the first working time completion over a planned time T. For each policy, the optimal schedule of preventive replacement that minimizes the mean cost rate is presented analytically and discussed numerically. Because the framework and analysis are general, the proposed models extend several existing results.     

A data-driven two-stage maintenance framework for degradation prediction in semiconductor manufacturing industries

To reduce the production costs and breakdown risks in industrial manufacturing systems, condition-based maintenance has been actively pursued for prediction of equipment degradation and optimization of maintenance schedules. In this paper, a two-stage maintenance framework using data-driven techniques under two training types will be developed to predict the degradation status in industrial applications. The proposed framework consists of three main blocks, namely, Primary Maintenance Block (PMB), Secondary Maintenance Block (SMB), and degradation status determination block. As the popular methods with deterministic training, back-propagation Neural Network (NN) and evolvable NN are employed in PMB for the degradation prediction. Another two data-driven methods with probabilistic training, namely, restricted Boltzmann machine and deep belief network are applied in SMB as the backup of PMB to model non-stationary processes with the complicated underlying characteristics. Finally, the multiple regression forecasting is adopted in both blocks to check prediction accuracies. The effectiveness of our proposed two-stage maintenance framework is testified with extensive computation and experimental studies on an industrial case of the wafer fabrication plant in semiconductor manufactories, achieving up to 74.1% in testing accuracies for equipment degradation prediction.     

Optimal filtering for networked systems with Markovian communication delays

This paper is concerned with optimal filter problems for networked systems with random transmission delays, while the delay process is modeled as a multi-state Markov chain. By defining a delay-free observation sequence, the optimal filter problems are transformed into ones of the Markov jumping parameter system. We first present an optimal Kalman filter, which is with time-varying, path-dependent filter gains, and the number of the paths grows exponentially in time delay. Thus an alternative optimal Markov jump linear filter is presented, in which the filter gains just depend on the present value of the Markov chain. Further, an optimal filter with constant-gains is developed, the existence condition for the stabilizing solutions to the filter is given, and it can be shown that the proposed Markov jump linear filter converges to the constant-gain filter under appropriate assumptions.     

Remote,condition-based maintenance for web-enabled robotic system

The current trends in industry include an integration of information and knowledge-base network with a manufacturing system, which coined a new term, e-manufacturing. From the perspective of e-manufacturing any production equipment and its control functions do not exist alone, instead becoming a part of the holistic operation system with distant monitoring, remote quality control, and fault diagnostic capabilities. The key to this new paradigm is the accessibility to a remotely located system and having the means of responding to a changing environment, which is better suited for today's rapidly changing environment. Within the framework of the web-enabled robotic system, this paper focuses on the remote maintenance schemes with an emphasis on condition-based maintenance strategies. Real-time monitoring of robot harmonic drive systems and operational status have been attained over the Web. A mathematical modeling of system availability has been derived in order to account for other failures that might occur in the subsystems of the robot. Compared to the schedule-based maintenance strategies, the proposed approach shows great potential for improving overall production efficiency, while reducing the cost of maintenance.     

Optimal maintenance policy incorporating system level and unit level for mechanical systems

The study works on a multi-level maintenance policy combining system level and unit level under soft and hard failure modes. The system experiences system-level preventive maintenance (SLPM) when the conditional reliability of entire system exceeds SLPM threshold, and also undergoes a two-level maintenance for each single unit, which is initiated when a single unit exceeds its preventive maintenance (PM) threshold, and the other is performed simultaneously the moment when any unit is going for maintenance. The units experience both periodic inspections and aperiodic inspections provided by failures of hard-type units. To model the practical situations, two types of economic dependence have been taken into account, which are set-up cost dependence and maintenance expertise dependence due to the same technology and tool/equipment can be utilised. The optimisation problem is formulated and solved in a semi-Markov decision process framework. The objective is to find the optimal system-level threshold and unit-level thresholds by minimising the long-run expected average cost per unit time. A formula for the mean residual life is derived for the proposed multi-level maintenance policy. The method is illustrated by a real case study of feed subsystem from a boring machine, and a comparison with other policies demonstrates the effectiveness of our approach.     

A condition-based maintenance policy and input parameters estimation for deteriorating systems under periodic inspection

This paper combines an optimization model and input parameters estimation from empirical data, in order to propose condition-based maintenance policies. The system deterioration is described by discrete states ordered from the state “as good as new” to the state “completely failed”. At each periodic inspection, whose outcome might not be accurate, a decision has to be made between continuing to operate the system or stopping and performing its preventive maintenance. We explore the problem of how to estimate the model input parameters, i.e., how to adequate the model inputs to the empirical data available. For this purpose, we use the Hidden Markov Model theory. The literature has not explored the combination of optimization techniques and model input parameters, through historical data, for problems with imperfect information such as the one considered in this paper. We thoroughly discuss our approach, illustrate it with empirical data and also point out directions for future research.     

Optimal pricing policies for services with consideration of facility maintenance costs

For survival and success, pricing is an essential issue for service firms. This article deals with the pricing strategies for services with substantial facility maintenance costs. For this purpose, a mathematical framework that incorporates service demand and facility deterioration is proposed to address the problem. The facility and customers constitute a service system driven by Poisson arrivals and exponential service times. A service demand with increasing price elasticity and a facility lifetime with strictly increasing failure rate are also adopted in modelling. By examining the bidirectional relationship between customer demand and facility deterioration in the profit model, the pricing policies of the service are investigated. Then analytical conditions of customer demand and facility lifetime are derived to achieve a unique optimal pricing policy. The comparative statics properties of the optimal policy are also explored. Finally, numerical examples are presented to illustrate the effects of parameter variations on the optimal pricing policy.     

Analysis of the TDLMS algorithm operating in a nonstationary environment

This paper presents a stochastic analysis of the transform-domain least-mean-square (TDLMS) algorithm operating in a nonstationary environment (time-varying plant) with real-valued correlated Gaussian input data, from which the analysis for the stationary environment follows as a special case. In this analysis, accurate model expressions are derived describing the transformed mean weight-error behavior, learning curve, transformed weight-error covariance matrix, steady-state excess mean-square error (EMSE), misadjustment, step size for minimum EMSE, degree of nonstationarity, as well as a relationship between misadjustment and degree of nonstationarity. Based on these model expressions, the impact of the algorithm parameters on its performance is discussed, clarifying the behavior of the algorithm vis-à-vis the nonstationary environment considered. Simulation results for the TDLMS algorithm are shown by using the discrete cosine transform, which confirm the accuracy of the proposed model for both transient and steady-state phases.     

Optimal spares and preventive maintenance frequencies for constrained industrial systems

The research investigates the influence of an effective maintenance system on the efficient performance of any industrial system. The core concept of the research explains that the simultaneous focus on the spares inventory subsystem as well as on the preventive maintenance subsystem must be considered when developing a quality maintenance programme. Considering and developing such aspects separately will lead to suboptimal performance since there exists a trade-off between overstocking and undersupplying spares for preventive maintenance activities. Details on the technique chosen are discussed, namely simulation modelling as well as recent developments such as agent based modelling. Advantages of this technique are the flexibility in representing complex relationships within a system without knowing the exact form. The optimisation heuristic, a genetic algorithm, which was used to solve the research problem is explained. Finally a case study is used to demonstrate the aptness and success of the research approach, namely an annual 44% maintenance cost saving and 3% increase in production output.     

考虑老化因素的串联系统不完全维修决策优化

针对Markov过程和虚拟役龄方法难以全面描述系统不完全维修的问题,构建基于准更新过程的串联系统可用度分析模型,提出平均故障间隔时间和平均维修间隔时间确定方法;引入运行时间缩减因子、维修时间增长因子描述考虑老化因素的不完全维修过程,确定系统可用度指标确定方法;根据单位时间维修成本、系统平均运行时间约束条件,确定系统在不同指标下的不完全维修次数.以某控制系统为例,应用传统方法验证不完全维修条件下可用度确定方法的有效性,并以维修成本、运行时间为约束确定不同条件下的不完全维修次数,为维修决策制定提供理论指导.     

Stochastic model predictive control for constrained discrete-time Markovian switching systems

In this paper we study constrained stochastic optimal control problems for Markovian switching systems, an extension of Markovian jump linear systems (MJLS), where the subsystems are allowed to be nonlinear. We develop appropriate notions of invariance and stability for such systems and provide terminal conditions for stochastic model predictive control (SMPC) that guarantee mean-square stability and robust constraint fulfillment of the Markovian switching system in closed-loop with the SMPC law under very weak assumptions. In the special but important case of constrained MJLS we present an algorithm for computing explicitly the SMPC control law off-line, that combines dynamic programming with parametric piecewise quadratic optimization.     

Joint optimization of preventive maintenance and inventory policies for multi-unit systems subject to deteriorating spare part inventory

The interconnection of maintenance and spare part inventory management often puzzles managers and researchers. The deterioration of the inventory affects decision-making and increases losses. Block replacement and periodic review inventory policies were here used to evaluate a joint optimization problem for multi-unit systems in the presence of inventory deterioration. The deterministic deteriorating inventory (DDI) model was used to describe deteriorating inventory when deteriorating inventory data were available and the stochastic deteriorating inventory (SDI) model was used when they were not. Analytical joint optimization models were established, and the preventive replacement interval and the maximum inventory level served as decision variables to minimize the expected system total cost rate. This work proved the existence of the optimal maximum inventory level and gave the uniqueness condition under the DDI model. Numerical experiments based on the electric locomotives in Slovenian Railways were performed to confirm the effectiveness of the proposed models. Results showed the total cost rate to be sensitive to the maximum inventory level, which indicates that the research of this work is necessary. Further, the optimal preventive replacement interval was reduced and the optimal maximum inventory level was increased to balance the influence of deteriorating inventory. Monte Carlo experiments were used to show that the proposed policy is better than policies that do not take deteriorating inventory into account.     

Machine learning for anomaly detection and process phase classification to improve safety and maintenance activities

Anomaly detection is a crucial aspect for both safety and efficiency of modern process industries.This paper proposes a two-steps methodology for anomaly detection in industrial processes, adopting machine learning classification algorithms. Starting from a real-time collection of process data, the first step identifies the ongoing process phase, the second step classifies the input data as “Expected”, “Warning”, or “Critical”. The proposed methodology is extremely relevant where machines carry out several operations without the evidence of production phases. In this context, the difficulty of attributing the real-time measurements to a specific production phase affects the success of the condition monitoring. The paper proposes the comparison of the anomaly detection step with and without the process phase identification step, validating its absolute necessity. The methodology applies the decision forests algorithm, as a well-known anomaly detector from industrial data, and decision jungle algorithm, never tested before in industrial applications. A real case study in the pharmaceutical industry validates the proposed anomaly detection methodology, using a 10 months database of 16 process parameters from a granulation process.     

An intelligent condition-based maintenance platform for rotating machinery

Maintenance is of necessity for sustaining machinery availability and reliability in order to ensure productivity, product quality, on-time delivery, and safe working environment. The costly maintenance strategies such as corrective maintenance and scheduled maintenance have been progressively replaced by superior maintenance strategies in which condition-based maintenance (CBM) is one of the delegates. This strategy commonly consists of sequent modules such as data acquisition, signal processing, feature extraction and feature selection, condition monitoring, etc. However, approaches in literature which have been developed for each module and implemented for different applications are standalone instead of a comprehensive system. Furthermore, these approaches have been demonstrated in a laboratory environment without any industrial validations. For these reasons, an intelligent algorithm based CBM platform is proposed in this paper to be applied for rotating machinery easily and effectively. Subsequently, two case-studies are presented in order to evaluate the effectiveness of this platform in industrial applications.     

Robust sampled-data control for Markovian jump linear systems

In this paper, we consider the problem of robust control for uncertain sampled-data systems that possess random jumping parameters which is described by a finite-state Markov process. The conditions for the existence of a stabilizing control and optimal control for the underlying systems are obtained. The desired controllers are designed which are in terms of matrix inequalities. Finally, a numerical example is given to show the potential of the proposed techniques.     

网络环境下计算机硬件的安全保障及维护策略

网络环境下计算机硬件面临新的安全问题,这需要提高对硬件安全保障及维护工作的重视度,提高计算机设备运行的安全质量。鉴于此,文章以网络环境为着手点,分析了硬件对计算机安全运行的影响及原理,总结了网络环境下计算机硬件面临的安全问题,并结合实际情况给出了做好硬件安全保障及维护的策略,希望进一步提高计算机设备运行的安全性。     

Intelligent workload balance control of the assembly process considering condition-based maintenance

Balancing the workloads of workstations is key to the efficiency of an assembly line. However, the initial balance can be broken by the changing processing abilities of machines because of machine degradation, and at some point, re-balancing of the line is inevitable. Nevertheless, the impacts of unexpected events on assembly line re-balancing are always ignored. With the advanced sensor technologies and Internet of Things, the machine degradation process can be monitored continuously, and condition-based maintenance can be implemented to improve the health state of each machine. With the technology of robotic process automation, workflows of the assembly process can be smoothed and workstations can work autonomously together. A higher level of process automation can be achieved. Real-time information of the processing abilities of machines will bring new opportunities for automated workload balance via adaptive decision-making. In this study, a fuzzy control system is developed to make real-time decisions to balance the workloads based on the processing abilities of workstations, given the policy of condition-based maintenance. Fuzzy controllers are used to decide whether to re-balance the assembly line and how to adjust the production rate of each workstation. The numerical experiments show that the buffer level of the assembly line with the proposed fuzzy control system is lower than that of the assembly line without any control system and the buffer level of the assembly line with another control system is the lowest. The demand can always be satisfied by assembly lines except the one with another control system since there is too much production loss sacrificed for the low buffer level. The sensitivity analysis of the control performance to the parameter settings is also conducted. Thus, the effectiveness of the proposed fuzzy control system is demonstrated, and intelligent automation can improve the performance of the assembly process by the fuzzy control system since real-time information of the assembly line can be used for adaptive decision-making.     

Optimal filtering for networked systems with stochastic sensor gain degradation

In this paper, the optimal filtering problem is investigated for a class of networked systems in the presence of stochastic sensor gain degradations. The degradations are described by sequences of random variables with known statistics. A new measurement model is put forward to account for sensor gain degradations, network-induced time delays as well as network-induced data dropouts. Based on the proposed new model, an optimal unbiased filter is designed that minimizes the filtering error variance at each time-step. The developed filtering algorithm is recursive and therefore suitable for online application. Moreover, both currently and previously received signals are utilized to estimate the current state in order to achieve a better accuracy. A numerical simulation is exploited to illustrate the effectiveness of the proposed algorithm.