Optimizing maintenance and repair policies via a combination of genetic algorithms and Monte Carlo simulation

In this paper we present an optimization approach based on the combination of a Genetic Algorithms maximization procedure with a Monte Carlo simulation. The approach is applied within the context of plant logistic management for what concerns the choice of maintenance and repair strategies. A stochastic model of plant operation is developed from the standpoint of its reliability/availability behavior, i.e. of the failure/repair/maintenance processes of its components. The model is evaluated by Monte Carlo simulation in terms of economic costs and revenues of operation. The flexibility of the Monte Carlo method allows us to include several practical aspects such as stand-by operation modes, deteriorating repairs, aging, sequences of periodic maintenances, number of repair teams available for different kinds of repair interventions (mechanical, electronic, hydraulic, etc.), components priority rankings. A genetic algorithm is then utilized to optimize the components maintenance periods and number of repair teams. The fitness function object of the optimization is a profit function which inherently accounts for the safety and economic performance of the plant and whose value is computed by the above Monte Carlo simulation model. For an efficient combination of Genetic Algorithms and Monte Carlo simulation, only few hundreds Monte Carlo histories are performed for each potential solution proposed by the genetic algorithm. Statistical significance of the results of the solutions of interest (i.e. the best ones) is then attained exploiting the fact that during the population evolution the fit chromosomes appear repeatedly many times. The proposed optimization approach is applied on two case studies of increasing complexity.     

A Monte Carlo simulation approach to the availability assessment of multi-state systems with operational dependencies

This paper deals with multi-state systems (MSS), whose performance can settle on different levels, e.g. 100%, 80%, 50% of the nominal capacity, depending on the operative conditions of the constitutive multi-state elements. Examples are manufacturing, production, power generation and gas and oil transportation systems. Often in practice, MSS are such that operational dependencies exist between the system state and the state of its components. For example, in a production line of nodal series structure, with no buffers between the nodes, if one of the nodes throughput changes (e.g. switches from 100% to 50% due to a deterministic or stochastic transition of one of its components), the other nodes must be reconfigured (i.e. their components must deterministically change their states) so as to provide the same throughput.In this paper, we present a Monte Carlo simulation technique which allows modelling the complex dynamics of multi-state components subject to operational dependencies with the system overall state. A correlation method is tailored to model the automatic change of state of the relevant components following a change in one of the system nodes. The proposed technique is verified on a simple case study of literature.     

Condition-based maintenance optimization by means of genetic algorithms and Monte Carlo simulation

Efficient maintenance policies are of fundamental importance in system engineering because of their fallbacks into the safety and economics of plants operation. When the condition of a system, such as its degradation level, can be continuously monitored, a Condition-Based Maintenance (CBM) policy can be implemented, according to which the decision of maintaining the system is taken dynamically on the basis of the observed condition of the system.In this paper, we consider a continuously monitored multi-component system and use a Genetic Algorithm (GA) for determining the optimal degradation level beyond which preventive maintenance has to be performed. The problem is framed as a multi-objective search aiming at simultaneously optimizing two typical objectives of interest, profit and availability. For a closer adherence to reality, the predictive model describing the evolution of the degrading system is based on the use of Monte Carlo (MC) simulation. More precisely, the flexibility offered by the simulation scheme is exploited to model the dynamics of a stress-dependent degradation process in load-sharing components and to account for limitations in the number of maintenance technicians available. The coupled (GA[plus ]MC) approach is rendered particularly efficient by the use of the ‘drop-by-drop’ technique, previously introduced by some of the authors, which allows to effectively drive the combinatorial search towards the most promising solutions.     

New method to minimize the preventive maintenance cost of series–parallel systems

General preventive maintenance model for input components of a system, which improves the reliability to ‘as good as new,’ was used to optimize the maintenance cost. The cost function of a maintenance policy was minimized under given availability constraint. An algorithm for first inspection vector of times was described and used on selected system example. A special ratio-criterion, based on the time dependent Birnbaum importance factor, was used to generate the ordered sequence of first inspection times. Basic system availability calculations of the paper were done by using simulation approach with parallel simulation algorithm for availability analysis. These calculations, based on direct Monte Carlo technique, were applied within the programming tool Matlab. A genetic algorithm optimization technique was used and briefly described to create the Matlab's algorithm to solve the problem of finding the best maintenance policy with a given restriction. Adjacent problem, which we called ‘reliability assurance,’ was also theoretically solved, concerning the increase of the cost when asymptotic availability value conforms to a given availability constraint.     

A rule induction approach to improve Monte Carlo system reliability assessment

A Decision Tree (DT) approach to build empirical models for use in Monte Carlo reliability evaluation is presented. The main idea is to develop an estimation algorithm, by training a model on a restricted data set, and replacing the Evaluation Function (EF) by a simpler calculation, which provides reasonably accurate model outputs. The proposed approach is illustrated with two systems of different size, represented by their equivalent networks. The robustness of the DT approach as an approximated method to replace the EF is also analysed. Excellent system reliability results are obtained by training a DT with a small amount of information.     

Genetic algorithms and Monte Carlo simulation for optimal plant design

We present an approach to the optimal plant design (choice of system layout and components) under conflicting safety and economic constraints, based upon the coupling of a Monte Carlo evaluation of plant operation with a Genetic Algorithms-maximization procedure. The Monte Carlo simulation model provides a flexible tool, which enables one to describe relevant aspects of plant design and operation, such as standby modes and deteriorating repairs, not easily captured by analytical models. The effects of deteriorating repairs are described by means of a modified Brown–Proschan model of imperfect repair which accounts for the possibility of an increased proneness to failure of a component after a repair. The transitions of a component from standby to active, and vice versa, are simulated using a multiplicative correlation model. The genetic algorithms procedure is demanded to optimize a profit function which accounts for the plant safety and economic performance and which is evaluated, for each possible design, by the above Monte Carlo simulation.In order to avoid an overwhelming use of computer time, for each potential solution proposed by the genetic algorithm, we perform only few hundreds Monte Carlo histories and, then, exploit the fact that during the genetic algorithm population evolution, the fit chromosomes appear repeatedly many times, so that the results for the solutions of interest (i.e. the best ones) attain statistical significance.     

A combination of Monte Carlo simulation and cellular automata for computing the availability of complex network systems

This paper combines Monte Carlo simulation and cellular automata for computing the availability of a complex network system and the importance measures of its elements.     

Direct simulation Monte Carlo modeling of metal vapor flows in application to thin film deposition

Thin films of metal for electronics, nano/microelectromechanical systems and optical coatings are often prepared by various vacuum deposition techniques. Modeling such metal vapor flows using methods such as the direct simulation Monte Carlo (DSMC) can aid in the design and analysis of deposition systems and accelerate development of films with desired properties. The determination of suitable variable hard sphere (VHS) molecular model parameters for DSMC simulations using measured growth rate distribution is demonstrated with aluminum vapor as an example. Axisymmetric DSMC simulations using a VHS model corresponding to a reference diameter of 0.8 nm and a viscosity-temperature exponent of 1 are shown to agree well with available experimental data. The model is then used in two-dimensional DSMC simulations to study the interaction of plumes from multiple sources. An expression for substrate mass flux assuming no interaction between sources agrees well with DSMC simulations for a mass flow rate of 0.1 g/min corresponding to a Knudsen number (Kn) of about 0.1. The non-additive interaction of plumes at a higher flow rate of 1 g/min corresponding to a Kn of about 0.01 results in a higher mass flux non-uniformity in the DSMC simulations which is not captured by the simplified analytical expression.     

Correlation control in small-sample Monte Carlo type simulations I: A simulated annealing approach

The objective of this paper is to propose an effective procedure for sampling from a multivariate population within the framework of Monte Carlo simulations. The typical application of the proposed approach involves a computer-based model, featuring random variables, in which it is impossible to find a way (closed form or numerical) to carry out the necessary transformation of the variables, and where simulation is expensive in terms of computing resources and time. Other applications of the proposed method can be seen in random field simulations, optimum learning sets for neural networks and response surfaces, and in the design of experiments.     

A Monte Carlo investigation of growth and characterization of heteroepitaxial thin films

We investigate the growth of mismatched thin films by a kinetic Monte Carlo computer simulation and including a local photoemission model with reflection high-energy electron diffraction (RHEED) intensity for comparison. The strain is introduced through an elastic energy term based on a valence force field approximation. We describe an atomistic mechanism for dislocation nucleation during first stage of GaSb/GaAs (001) growth and in situ variations of photoemission current (PE) and RHEED intensity are reported. We have shown the formation of grooves corresponding to (111) facets, a precursor to the formation of misfit defects. The surface roughening and facetting by creation of grooves explain the absence of photoemission and RHEED oscillations in accordance with experimental observations [J.J. Zinck and D.H. Chow, J. Cryst. Growth, 175/176 (1997) 323, J.J. Zinck and D.H Chow, Appl. Phys. Lett. 66 (1995) 3524].     

Modeling uncertainty in risk assessment: An integrated approach with fuzzy set theory and Monte Carlo simulation

Modeling uncertainty during risk assessment is a vital component for effective decision making. Unfortunately, most of the risk assessment studies suffer from uncertainty analysis. The development of tools and techniques for capturing uncertainty in risk assessment is ongoing and there has been a substantial growth in this respect in health risk assessment. In this study, the cross-disciplinary approaches for uncertainty analyses are identified and a modified approach suitable for industrial safety risk assessment is proposed using fuzzy set theory and Monte Carlo simulation. The proposed method is applied to a benzene extraction unit (BEU) of a chemical plant. The case study results show that the proposed method provides better measure of uncertainty than the existing methods as unlike traditional risk analysis method this approach takes into account both variability and uncertainty of information into risk calculation, and instead of a single risk value this approach provides interval value of risk values for a given percentile of risk. The implications of these results in terms of risk control and regulatory compliances are also discussed.     

A single scatter electron Monte Carlo approach for simulating gamma-ray stopping efficiencies of Geiger-Müller counters

In spite of their relatively poor gamma-ray stopping efficiencies, the Geiger-Müller (GM) counter is still preferred in many radioisotope gauges for industrial measurements. This is because these detectors exhibit a high degree of robustness in harsh environments, are relatively insensitive to temperature changes in the environment, and are inexpensive compared to other types of radiation detectors. These properties could make the use of GM counters very feasible in a number of industrial applications, such as gamma-ray tomography and gamma-ray density gauges, provided that their gamma-ray stopping efficiencies can be improved. The Monte Carlo (MC) method is a powerful computational physics tool that is utilized very often in the design of radiation detectors and radioisotope gauges. In this work a MC model for GM counters that is benchmarked with experiments at the primary photon energy of 59.5 keV is proposed. This is a specific purpose MC simulation code that, as opposed to publicly available general purpose MC codes, employs single scatter (or microscopic) electron transport and is currently under development. In this paper, the MC code is described in detail and the results of the specific purpose MC code are benchmarked with experiments and two general purpose MC codes, MCNP5 and PENELOPE. It was observed that the specific purpose MC code improved the reduced chi-square value when compared to MCNP5 and PENELOPE.     

A reliability/availability approach to joint production and maintenance scheduling with multiple preventive maintenance services

In classical scheduling problems, it is often assumed that the machines are available during the whole planning horizon, while in realistic environments, machines need to be maintained and therefore may become unavailable within production periods. Hence, in this paper we suggest a joint production and maintenance scheduling (JPMS) with multiple preventive maintenance services, in which the reliability/availability approach is employed to model the maintenance aspects of a problem. To cope with the suggested JPMS, a mixed integer nonlinear programming model is developed and then a population-based variable neighbourhood search (PVNS) algorithm is devised for a solution method. In order to enhance the search diversification of basic variable neighbourhood search (VNS), our PVNS uses an epitome-based mechanism in each iteration to transform a group of initial individuals into a new solution, and then multiple trial solutions are generated in the shaking stage for a given solution. At the end of the local search stage, the best obtained solution by all of the trial solutions is recorded and the worst solution in population is replaced with this new solution. The evolution procedure is continued until a predefined number of iterations is violated. To validate the effectiveness and robustness of PVNS, an extensive computational study is implemented and the simulation results reveal that our PVNS performs better than traditional algorithms, especially in large size problems.     

A semi-analytical Monte Carlo simulation method for system's reliability with load sharing and damage accumulation

A novel approach for assessing a systems' reliability with dependency structures, load sharing, and damage accumulation and reversal is proposed in this paper. It is a blend of analytical reliability analysis performed at the component level, and is based on understanding the failure mechanism of the components, and a Monte Carlo simulation for the entire system to assess the reliability at the system level incorporating the dynamics of the system behavior as the components interact, share loads, and age over time. Model reduction is deployed to reduce the complexity and accelerate the simulation and convergence of the analytical methods such as FORM and SORM performed at the component level. Numerical examples are provided to illustrate the usability and performance of the method.     

A Bayesian approach to modeling and predicting pitting flaws in steam generator tubes

Steam generators in nuclear power plants have experienced varying degrees of under-deposit pitting corrosion. A probabilistic model to accurately predict pitting damage is necessary for effective life-cycle management of steam generators. This paper presents an advanced probabilistic model of pitting corrosion characterizing the inherent randomness of the pitting process and measurement uncertainties of the in-service inspection (ISI) data obtained from eddy current (EC) inspections. A Markov chain Monte Carlo simulation-based Bayesian method, enhanced by a data augmentation technique, is developed for estimating the model parameters. The proposed model is able to predict the actual pit number, the actual pit depth as well as the maximum pit depth, which is the main interest of the pitting corrosion model. The study also reveals the significance of inspection uncertainties in the modeling of pitting flaws using the ISI data: Without considering the probability-of-detection issues and measurement errors, the leakage risk resulted from the pitting corrosion would be under-estimated, despite the fact that the actual pit depth would usually be over-estimated.     

Squeeze-and-breathe evolutionary Monte Carlo optimization with local search acceleration and its application to parameter fitting

Estimating parameters from data is a key stage of the modelling process, particularly in biological systems where many parameters need to be estimated from sparse and noisy datasets. Over the years, a variety of heuristics have been proposed to solve this complex optimization problem, with good results in some cases yet with limitations in the biological setting. In this work, we develop an algorithm for model parameter fitting that combines ideas from evolutionary algorithms, sequential Monte Carlo and direct search optimization. Our method performs well even when the order of magnitude and/or the range of the parameters is unknown. The method refines iteratively a sequence of parameter distributions through local optimization combined with partial resampling from a historical prior defined over the support of all previous iterations. We exemplify our method with biological models using both simulated and real experimental data and estimate the parameters efficiently even in the absence of a priori knowledge about the parameters.     

Simulation at high temperature of atomic deposition, islands coalescence, Ostwald and inverse Ostwald ripening with a general simple kinetic Monte Carlo code

А three-dimensional kinetic Monte Carlo model (kMC) is proposed for the simulation of deposition and evolution of surface structures at elevated temperatures. The code includes deposition of one given type of atom and main thermally driven events such as surface diffusion, diffusion along island edges, detachment from islands, and movement of atoms on deposited surfaces. It can be used not only for simulating nucleation and growth of thin films but also for simulating time evolution of a given structure when annealed. It is a specific event kMC code, and the rates of the events are used as inputs. It allows the simulation of thousands of incident particles and the simulation of a system at high temperature without suffering large computational time. The code runs on a PC and is freely available.Results of modeling various situations like atomic deposition (Pd on SiO₂), islands coalescence (Cu on Cu), Ostwald and inverse Ostwald ripening (Co/C and Co/SiO₂) were tested against existing experimental and theoretical data and show a good agreement for all those cases.     

A tolerance interval based approach to address uncertainty for RAMS+C optimization

This paper proposes an approach based on tolerance intervals to address uncertainty for RAMS+C informed optimization of design and maintenance of safety-related systems using a combined Monte Carlo (MC) (simulation) and Genetic Algorithm (search) procedure. This approach is intended to keep control of the uncertainty effects on the decision criteria and reduce the computational effort in simulating RAMS+C using a MC procedure with simple random sampling. It exploits the advantages of order statistics to provide distribution free tolerance intervals for the RAMS+C estimation, which is based on the minimum number of runs necessary to guarantee a probability content or coverage with a confidence level. This approach has been implemented into a customization of the Multi-Objective Genetic Algorithm introduced by the authors in a previous work. For validation purposes, a simple application example regarding the testing and maintenance optimization of the High-Pressure Injection System of a nuclear power plant is also provided, which considers the effect of the epistemic uncertainty associated with the equipment reliability characteristics on the optimal testing and maintenance policy. This example proves that the new approach can provide a robust, fast and powerful tool for RAMS+C informed multi-objective optimization of testing and maintenance under uncertainty in objective and constraints. It is shown that the approach proposed performs very favourably in the face of noise in the output (i.e. uncertainty) and it is able to find the optimum over a complicated, high-dimensional non-linear space in a tiny fraction of the time required for enumeration of the decision space. In addition, a sensitivity study on the number of generations versus the number of trials (i.e. simulation runs) shows that overall computational resources must be assigned preferably to evolving a larger number of generations instead of being more precise in the quantification of the RAMS+C attributes for a candidate solution, i.e. evolution is preferred to accuracy.