Högskolan i Skövde

This study investigates the advantages and potentials of the metamodelbased multi-objective optimization (MOO) of a turning operation through the application of finite element simulations and evolutionary algorithms to a metal cutting process. The objectives are minimizing the interface temperature and tool wear depth obtained from FE simulations using DEFORM2D software, and maximizing the material removal rate. Tool geometry and process parameters are considered as the input variables. Seven metamodelling methods are employed and evaluated, based on accuracy and suitability. Radial basis functions with a priori bias and Kriging are chosen to model tool–chip interface temperature and tool wear depth, respectively. The non-dominated solutions are found using the strength Pareto evolutionary algorithm SPEA2 and compared with the non-dominated front obtained from pure simulation-based MOO. The metamodel-based MOO method is not only advantageous in terms of reducing the computational time by 70%, but is also able to discover 31 new non-dominated solutions over simulation-based MOO.

This study investigates the advantages and potentials of the metamodelbased multi-objective optimization (MOO) of a turning operation through the application of finite element simulations and evolutionary algorithms to a metal cutting process. The objectives are minimizing the interface temperature and tool wear depth obtained from FE simulations using DEFORM2D software, and maximizing the material removal rate. Tool geometry and process parameters are considered as the input variables. Seven metamodelling methods are employed and evaluated, based on accuracy and suitability. Radial basis functions with a priori bias and Kriging are chosen to model tool–chip interface temperature and tool wear depth, respectively. The non-dominated solutions are found using the strength Pareto evolutionary algorithm SPEA2 and compared with the non-dominated front obtained from pure simulation-based MOO. The metamodel-based MOO method is not only advantageous in terms of reducing the computational time by 70%, but is also able to discover 31 new non-dominated solutions over simulation-based MOO.

Radial basis functions are augmented with a posteriori bias in order to perform robustly when used as metamodels. Recently, it has been proposed that the bias can simply be set a priori by using the normal equation, i.e., the bias becomes the corresponding regression model. In this study, we demonstrate the performance of the suggested approach (RBFpri) with four other well-known metamodeling methods; Kriging, support vector regression, neural network and multivariate adaptive regression. The performance of the five methods is investigated by a comparative study, using 19 mathematical test functions, with five different degrees of dimensionality and sampling size for each function. The performance is evaluated by root mean squared error representing the accuracy, rank error representing the suitability of metamodels when coupled with evolutionary optimization algorithms, training time representing the efficiency and variation of root mean squared error representing the robustness. Furthermore, a rigorous statistical analysis of performance metrics is performed. The results show that the proposed radial basis function with a priori bias achieved the best performance in most of the experiments in terms of all three metrics. When considering the statistical analysis results, the proposed approach again behaved the best, while Kriging was relatively as accurate and support vector regression was almost as fast as RBFpri. The proposed RBF is proven to be the most suitable method in predicting the ranking among pairs of solutions utilized in evolutionary algorithms. Finally, the comparison study is carried out on a real-world engineering optimization problem.

The use of optimization to solve a simulation-based multi-objective problem produces a set of solutions that provide information about the trade-offs that have to be considered by the decision maker. An incomplete or sub-optimal set of solutions will negatively affect the quality of any subsequent decisions. The parameters that control the search behavior of an optimization algorithm can be used to minimize this risk. However, choosing good parameter settings for a given optimization algorithm and problem combination is difficult. The aim of this paper is to take a step towards optimal parameter settings for optimization of simulation-based problems. Two parameter tuning methods, Latin Hypercube Sampling and Genetic Algorithms, are used to maximize the performance of NSGA-II applied to a simulation-based problem with discrete variables. The strengths and weaknesses of both methods are analyzed. The effect of the number of decision variables and the function budget on the optimal parameter settings is also studied.

Evolutionary optimization algorithms typically use one or more parameters that control their behavior. These parameters, which are often kept constant, can be tuned to improve the performance of the algorithm on specific problems. However, past studies have indicated that the performance can be further improved by adapting the parameters during runtime. A limitation of these studies is that they only control, at most, a few parameters, thereby missing potentially beneficial interactions between them. Instead of finding a direct control mechanism, the novel approach in this paper is to use different parameter sets in different stages of an optimization. These multiple parameter sets, which remain static within each stage, are tuned through extensive bi-level optimization experiments that approximate the optimal adaptation of the parameters. The algorithmic performance obtained with tuned multiple parameter sets is compared against that obtained with a single parameter set. For the experiments in this paper, the parameters of NSGA-II are tuned when applied to the ZDT, DTLZ and WFG test problems. The results show that using multiple parameter sets can significantly increase the performance over a single parameter set.

Evolutionary optimization algorithms have parameters that are used to adapt the search strategy to suit different optimization problems. Selecting the optimal parameter values for a given problem is difficult without a-priori knowledge. Experimental studies can provide this knowledge by finding the best parameter values for a specific set of problems. This knowledge can also be constructed into heuristics (rule-of-thumbs) that can adapt the parameters for the problem. The aim of this paper is to assess the heuristics of the Covariance Matrix Adaptation Evolution Strategy (CMA-ES) optimization algorithm. This is accomplished by tuning CMA-ES parameters so as to maximize its performance on the CEC'15 problems, using a bilevel optimization approach that searches for the optimal parameter values. The optimized parameter values are compared against the parameter values suggested by the heuristics. The difference between specialized and generalized parameter values are also investigated.

The performance of an Evolutionary Algorithm (EA) can be greatly influenced by its parameters. The optimal parameter settings are also not necessarily the same across different problems. Finding the optimal set of parameters is therefore a difficult and often time-consuming task. This paper presents results of parameter tuning experiments on the NSGA-II and NSGA-III algorithms using the ZDT test problems. The aim is to gain new insights on the characteristics of the optimal parameter settings and to study if the parameters impose the same effect on both NSGA-II and NSGA-III. The experiments also aim at testing if the rule of thumb that the mutation probability should be set to one divided by the number of decision variables is a good heuristic on the ZDT problems. A comparison of the performance of NSGA-II and NSGA-III on the ZDT problems is also made.

Most optimization algorithms extract important algorithmic design decisions as control parameters. This is necessary because different problems can require different search strategies to be solved effectively. The control parameters allow for the optimization algorithm to be adapted to the problem at hand. It is however difficult to predict what the optimal control parameters are for any given problem. Finding these optimal control parameter values is referred to as the parameter tuning problem. One approach of solving the parameter tuning problem is to use a bilevel optimization where the parameter tuning problem itself is formulated as an optimization problem involving algorithmic performance as the objective(s). In this paper, we present a framework and architecture that can be used to solve large-scale parameter tuning problems using a bilevel optimization approach. The proposed framework is used to show that evolutionary algorithms are competitive as tuners against irace which is a state-of-the-art tuning method. Two evolutionary algorithms, differential evaluation (DE) and a genetic algorithm (GA) are evaluated as tuner algorithms using the proposed framework and software architecture. The importance of replicating optimizations and avoiding local optima is also investigated. The architecture is deployed and tested by running millions of optimizations using a computing cluster. The results indicate that the evolutionary algorithms can consistently find better control parameter values than irace. The GA, however, needs to be configured for an explicit exploration and exploitation strategy in order avoid local optima.

We consider a bilevel parameter tuning problem where the goal is to maximize the performance of a given multi-objective evolutionary optimizer on a given problem. The search for optimal algorithmic parameters requires the assessment of several sets of parameters, through multiple optimization runs, in order to mitigate the effect of noise that is inherent to evolutionary algorithms. This task is computationally expensive and therefore, in this paper, we propose to use sampling and metamodeling to approximate the performance of the optimizer as a function of its parameters. While such an approach is not unheard of, the choice of the metamodel to be used still remains unclear. The aim of this paper is to empirically compare 11 different metamodeling techniques with respect to their accuracy and training times in predicting two popular multi-objective performance metrics, namely, the hypervolume and the inverted generational distance. For the experiments in this pilot study, NSGA-II is used as the multi-objective optimizer for solving ZDT problems, 1 through 4.

This paper generalizes the automated innovization framework using genetic programming in the context of higher-level innovization. Automated innovization is an unsupervised machine learning technique that can automatically extract significant mathematical relationships from Pareto-optimal solution sets. These resulting relationships describe the conditions for Pareto-optimality for the multi-objective problem under consideration and can be used by scientists and practitioners as thumb rules to understand the problem better and to innovate new problem solving techniques; hence the name innovization (innovation through optimization). Higher-level innovization involves performing automated innovization on multiple Pareto-optimal solution sets obtained by varying one or more problem parameters. The automated innovization framework was recently updated using genetic programming. We extend this generalization to perform higher-level automated innovization and demonstrate the methodology on a standard two-bar bi-objective truss design problem. The procedure is then applied to a classic case of inventory management with multi-objective optimization performed at both system and process levels. The applicability of automated innovization to this area should motivate its use in other avenues of operational research.

Multi-objective optimization yields multiple solutions each of which is no better or worse than the others when the objectives are conflicting. These solutions lie on the Pareto-optimal front which is a lower-dimensional slice of the objective space. Together, the solutions may possess special properties that make them optimal over other feasible solutions. Innovization is the process of extracting such special properties (or design principles) from a trade-off dataset in the form of mathematical relationships between the variables and objective functions. In this paper, we deal with a closely related concept called temporal innovization. While innovization concerns the design principles obtained from the trade-off front, temporal innovization refers to the evolution of these design principles during the optimization process. Our study indicates that not only do different design principles evolve at different rates, but that they start evolving at different times. We illustrate temporal innovization using several examples.

Consumer-oriented companies are getting increasingly more sensitive about customer's perception of their products, not only to get a feedback on their popularity, but also to improve the quality and service through a better understanding of design issues for further development. However, a consumer's perception is often qualitative and is achieved through third party surveys or the company's recording of after-sale feedback through explicit surveys or warranty based commitments. In this paper, we consider an automobile company's warranty records for different vehicle models and suggest a data mining procedure to assign a customer satisfaction index (CSI) to each vehicle model based on the perceived notion of the level of satisfaction of customers. Based on the developed CSI function, customers are then divided into satisfied and dissatisfied customer groups. The warranty data are then clustered separately for each group and analyzed to find possible causes (field failures) and their relative effects on customer's satisfaction (or dissatisfaction) for a vehicle model. Finally, speculative introspection has been made to identify the amount of improvement in CSI that can be achieved by the reduction of some critical field failures through better design practices. Thus, this paper shows how warranty data from customers can be utilized to have a better perception of ranking of a product compared to its competitors in the market and also to identify possible causes for making some customers dissatisfied and eventually to help percolate these issues at the design level. This closes the design cycle loop in which after a design is converted into a product, its perceived level of satisfaction by customers can also provide valuable information to help make the design better in an iterative manner. The proposed methodology is generic and novel, and can be applied to other consumer products as well.

Multi-objective evolutionary algorithms (MOEAs)are often criticized for their high-computational costs. Thisbecomes especially relevant in simulation-based optimizationwhere the objectives lack a closed form and are expensive toevaluate. Over the years, meta-modeling or surrogate modelingtechniques have been used to build inexpensive approximationsof the objective functions which reduce the overall number offunction evaluations (simulations). Some recent studies however,have pointed out that accurate models of the objective functionsmay not be required at all since evolutionary algorithms onlyrely on the relative ranking of candidate solutions. Extendingthis notion to MOEAs, algorithms which can ‘learn’ Paretodominancerelations can be used to compare candidate solutionsunder multiple objectives. With this goal in mind, in thispaper, we study the performance of ten different off-the-shelfclassification algorithms for learning Pareto-dominance relationsin the ZDT test suite of benchmark problems. We considerprediction accuracy and training time as performance measureswith respect to dimensionality and skewness of the training data.Being a preliminary study, this paper does not include results ofintegrating the classifiers into the search process of MOEAs.

Metamodeling plays an important role in simulation-based optimization by providing computationally inexpensive approximations for the objective and constraint functions. Additionally metamodeling can also serve to filter noise, which is inherent in many simulation problems causing optimization algorithms to be mislead. In this paper, we conduct a thorough statistical comparison of four popular metamodeling methods with respect to their approximation accuracy at various levels of noise. We use six scalable benchmark problems from the optimization literature as our test suite. The problems have been chosen to represent different types of fitness landscapes, namely, bowl-shaped, valley-shaped, steep ridges and multi-modal, all of which can significantly influence the impact of noise. Each metamodeling technique is used in combination with four different noise handling techniques that are commonly employed by practitioners in the field of simulation-based optimization. The goal is to identify the metamodeling strategy, i.e. a combination of metamodeling and noise handling, that performs significantly better than others on the fitness landscapes under consideration. We also demonstrate how these results carry over to a simulation-based optimization problem concerning a scalable discrete event model of a simple but realistic production line.

Optimization of production systems often involves numerous simulations of computationally expensive discrete-event models. When derivative-free optimization is sought, one usually resorts to evolutionary and other population-based meta-heuristics. These algorithms typically demand a large number of objective function evaluations, which in turn, drastically increases the computational cost of simulations. To counteract this, meta-models are used to replace expensive simulations with inexpensive approximations. Despite their widespread use, a thorough evaluation of meta-modeling methods has not been carried out yet to the authors' knowledge. In this paper, we analyze 10 different meta-models with respect to their accuracy and training time as a function of the number of training samples and the problem dimension. For our experiments, we choose a standard discrete-event model of an unpaced flow line with scalable number of machines and buffers. The best performing meta-model is then used with an evolutionary algorithm to perform multi-objective optimization of the production model.

Practical multi-objective optimization problems often involve several decision variables that influence the objective space in different ways. All variables may not be equally important in determining the trade-offs of the problem. Decision makers, who are usually only concerned with the objective space, have a hard time identifying such important variables and understanding how the variables impact their decisions and vice versa. Several graphical methods exist in the MCDM literature that can aid decision makers in visualizing and navigating high-dimensional objective spaces. However, visualization methods that can specifically reveal the relationship between decision and objective space have not been developed so far. We address this issue through a novel visualization technique called trend mining that enables a decision maker to quickly comprehend the effect of variables on the structure of the objective space and easily discover interesting variable trends. The method uses moving averages with different windows to calculate an interestingness score for each variable along predefined reference directions. These scores are presented to the user in the form of an interactive heatmap. We demonstrate the working of the method and its usefulness through a benchmark and two engineering problems.

Real-world optimization problems typically involve multiple objectives to be optimized simultaneously under multiple constraints and with respect to several variables. While multi-objective optimization itself can be a challenging task, equally difficult is the ability to make sense of the obtained solutions. In this two-part paper, we deal with data mining methods that can be applied to extract knowledge about multi-objective optimization problems from the solutions generated during optimization. This knowledge is expected to provide deeper insights about the problem to the decision maker, in addition to assisting the optimization process in future design iterations through an expert system. The current paper surveys several existing data mining methods and classifies them by methodology and type of knowledge discovered. Most of these methods come from the domain of exploratory data analysis and can be applied to any multivariate data. We specifically look at methods that can generate explicit knowledge in a machine-usable form. A framework for knowledge-driven optimization is proposed, which involves both online and offline elements of knowledge discovery. One of the conclusions of this survey is that while there are a number of data mining methods that can deal with data involving continuous variables, only a few ad hoc methods exist that can provide explicit knowledge when the variables involved are of a discrete nature. Part B of this paper proposes new techniques that can be used with such datasets and applies them to discrete variable multi-objective problems related to production systems. 

The first part of this paper served as a comprehensive survey of data mining methods that have been used to extract knowledge from solutions generated during multi-objective optimization. The current paper addresses three major shortcomings of existing methods, namely, lack of interactiveness in the objective space, inability to handle discrete variables and inability to generate explicit knowledge. Four data mining methods are developed that can discover knowledge in the decision space and visualize it in the objective space. These methods are (i) sequential pattern mining, (ii) clustering-based classification trees, (iii) hybrid learning, and (iv) flexible pattern mining. Each method uses a unique learning strategy to generate explicit knowledge in the form of patterns, decision rules and unsupervised rules. The methods are also capable of taking the decision maker's preferences into account to generate knowledge unique to preferred regions of the objective space. Three realistic production systems involving different types of discrete variables are chosen as application studies. A multi-objective optimization problem is formulated for each system and solved using NSGA-II to generate the optimization datasets. Next, all four methods are applied to each dataset. In each application, the methods discover similar knowledge for specified regions of the objective space. Overall, the unsupervised rules generated by flexible pattern mining are found to be the most consistent, whereas the supervised rules from classification trees are the most sensitive to user-preferences. 

Most preference-based multi-objective evolutionary algorithms use reference points to articulate the decision maker's preferences. Since these algorithms typically converge to a sub-region of the Pareto-optimal front, the use of conventional performance measures (such as hypervolume and inverted generational distance) may lead to misleading results. Therefore, experimental studies in preference-based optimization often resort to using graphical methods to compare various algorithms. Though a few ad-hoc measures have been proposed in the literature, they either fail to generalize or involve parameters that are non-intuitive for a decision maker. In this paper, we propose a performance metric that is simple to implement, inexpensive to compute, and most importantly, does not involve any parameters. The so called expanding hypercube metric has been designed to extend the concepts of convergence and diversity to preference optimization. We demonstrate its effectiveness through constructed preference solution sets in two and three objectives. The proposed metric is then used to compare two popular reference-point based evolutionary algorithms on benchmark optimization problems up to 20 objectives.

Due to the nature of today's manufacturing industry, where enterprises are subjected to frequent changes and volatile markets, reconfigurable manufacturing systems (RMS) are crucial when addressing ramp-up and ramp-down scenarios derived from, among other challenges, increasingly shortened product lifecycles. Applying simulation-based optimization techniques to their designs under different production volume scenarios has become valuable when RMS becomes more complex. Apart from proposing the optimal solutions subject to various production volume changes, decision-makers can extract propositional knowledge to better understand the RMS design and support their decision-making through a knowledge discovery method by combining simulation-based optimization and data mining techniques. In particular, this study applies a novel flexible pattern mining algorithm to conduct post-optimality analysis on multi-dimensional, multi-objective optimization datasets from an industrial-inspired application to discover the rules regarding how the tasks are assigned to the workstations constitute reasonable solutions for scalable RMS. 

By innovatively formulating a bottleneck identication problem into a bi-objective optimization,simulation-based multi-objective optimization (SMO) can be eectively used as a new method for gen-eral production systems improvement. In a single optimization run, all attainable, maximum throughputlevels of the system can be sought through various optimal combinations of improvement changes ofthe resources. Additionally, the post-optimality frequency analysis on the Pareto-optimal solutions cangenerate a rank order of the attributes of the resources required to achieve the target throughput levels.Observing that existing research mainly put emphasis on measuring the convergence of the optimizationin the objective space, leaving no information on when the solutions in the decision space have convergedand stabilized, this paper represents the rst eort in increasing the knowledge about the convergence ofSMO for the rank ordering in the context of bottleneck analysis. By customizing the Spearman's footruledistance and Kendall's tau, this paper presents how these metrics can be used eectively to provide thedesired visual aid in determining the convergence of bottleneck ranking, hence can assist the user todetermine correctly the terminating condition of the optimization process. It illustrates and evaluatesthe convergence of the SMO for bottleneck analysis on a set of scalable benchmark models as well as twoindustrial simulation models. The results have shed promising direction of applying these new metrics tocomplex, real-world applications.

Bottleneck analysis can be defined as the process that includes both bottleneck identification and improvement. In the literature most of the proposed bottleneck-related methods address mainly bottleneck detection. By innovatively formulating a bottleneck analysis into a bi-objective optimization method, recent research has shown that all attainable, maximized TH of a production system, through various combinations of improvement changes of the resources, can be sought in a single optimization run. Nevertheless, when applied to simulation-based evaluation, such a bi-objective optimization is computationally expensive especially when the simulation model is complex and/or with a large amount of decision variables representing the improvement actions. The aim of this paper is therefore to introduce a novel variables screening enabled bi-objective optimization that is customized for bottleneck analysis of production systems. By using the Sequential Bifurcation screening technique which is particularly suitable for large-scale simulation models, fewer simulation runs are required to find the most influenacing factors in a simulation model. With the knowledge of these input variables, the bi-objective optimization used in the bottleneck analysis can customize the genetic operators on these variables individually according to their rank of main effects with the target to speed up the entire optimization process. The screening-enabled algorithm is then applied to a set of experiments designed to evaluate how well it performs when the number of variables increases is a scalable, benchmark model, as well as two real-world industrial-scale simulation models found in the automotive industry. The results have illustrated the promising direction of incorporating the knowledge of influencing variables and variable-wise genetic operators into a multi-objective optimization algorithm for bottleneck analysis.

Most evolutionary many-objective optimization (EMaO) algorithms start with a description of a number of the predefined set of reference points on a unit simplex. So far, most studies have used the Das and Dennis's structured approach for generating well-spaced reference points. Due to the highly structured nature of the procedure, this method cannot produce an arbitrary number of points, which is desired in an EMaO application. Although a layer-wise implementation has been suggested, EMO researchers always felt the need for a more generic approach. Motivated by earlier studies, we introduce a metric for defining well-spaced points on a unit simplex and propose a number of viable methods for generating such a set. We compare the proposed methods on a variety of performance metrics such as hypervolume (HV), deviation in triangularized simplices, distance of the closest point pair, and variance of the geometric means to nearest neighbors in up to 15-D spaces. We show that an iterative improvement based on Riesz s-energy is able to effectively find an arbitrary number of well-spaced points even in higher-dimensional spaces. Reference points created using the proposed Riesz s-energy method for a number of standard combinations of objectives and reference points as well as a source code written in Python are available publicly at https://www.egr.msu.edu/coinlab/blankjul/uniform. © 1997-2012 IEEE.

Computational optimization methods are most often used to find a single or multiple optimal or near-optimal solutions to the underlying optimization problem describing the problem at hand. In this paper, we elevate the use of optimization to a higher level in arriving at useful problem knowledge associated with the optimal or near-optimal solutions to a problem. In the proposed innovization process, first a set of trade-off optimal or near-optimal solutions are found using an evolutionary algorithm. Thereafter, the trade-off solutions are analyzed to decipher useful relationships among problem entities automatically so as to provide a better understanding of the problem to a designer or a practitioner. We provide an integrated algorithm for the innovization process and demonstrate the usefulness of the procedure to three real-world engineering design problems. New and innovative design principles obtained in each case should clearly motivate engineers and practitioners for its further application to more complex problems and its further development as a more efficient data analysis procedure.

Most of the recently proposed evolutionary many-objective optimization (EMO) algorithms start with a number of predefined reference points on a unit simplex. These algorithms use reference points to create reference directions in the original objective space and attempt to find a single representative near Pareto-optimal point around each direction. So far, most studies have used Das and Dennis’s structured approach for generating a uniformly distributed set of reference points on the unit simplex. Due to the highly structured nature of the procedure, this method does not scale well with an increasing number of objectives. In higher dimensions, most created points lie on the boundary of the unit simplex except for a few interior exceptions. Although a level-wise implementation of Das and Dennis’s approach has been suggested, EMO researchers always felt the need for a more generic approach in which any arbitrary number of uniformly distributed reference points can be created easily at the start of an EMO run. In this paper, we discuss a number of methods for generating such points and demonstrate their ability to distribute points uniformly in 3 to 15-dimensional objective spaces.

Modern decision support systems need to be connected online to equipment so that the large amount of data available can be used to guide the decisions of shop floor operators, making full use of the potential of industrial manufacturing systems. This paper investigates a novel optimization and data analytic method to implement such a decision support system, based on heuristic generation using genetic programming and simulation-based optimization running on a digital twin. Such a digital-twin-based decision support system allows the proactively searching of the best attribute combinations to be used in a data-driven composite dispatching rule for the short-term corrective maintenance task prioritization. Both the job (e.g., bottlenecks) and operator priorities use multiple criteria, including competence, utilization, operator walking distances on the shop floor, bottlenecks, work-in-process, and parallel resource availability. The data-driven composite dispatching rules are evaluated using a digital twin, built for a real-world machining line, which simulates the effects of decisions regarding disruptions. Experimental results show improved productivity because of using the composite dispatching rules generated by such heuristic generation method compared to the priority dispatching rules based on similar attributes and methods. The improvement is more pronounced when the number of operators is reduced. This paper thus offers new insights about how shop floor data can be transformed into useful knowledge with a digital-twin-based decision support system to enhance resource efficiency.

The integration of simulation-based optimization and data mining is an emerging approach to support decision-making in the design and improvement of manufacturing systems. In such an approach, knowledge extracted from the optimal solutions generated by the simulation-based optimization process can provide important information to decision makers, such as the importance of the decision variables and their influence on the design objectives, which cannot easily be obtained by other means. However, can the extracted knowledge be directly used during the optimization process to further enhance the quality of the solutions? This paper proposes such an online knowledge extraction approach that is used together with a preference-guided multi-objective optimization algorithm on simulation models of manufacturing systems. Specifically, it introduces a combination of the multi-objective evolutionary optimization algorithm, NSGA-II, and a customized data mining algorithm, called Flexible Pattern Mining (FPM), which can extract knowledge in the form of rules in an online and automatic manner, in order to guide the optimization to converge towards a decision maker's preferred region in the objective space. Through a set of application problems, this paper demonstrates how the proposed FPM-NSGA-II can be used to support higher quality decision-making in manufacturing.

This paper describes a decision support system (DSS) built on knowledge extraction using simulation-based optimization and data mining. The paper starts with a requirements analysis based on a survey conducted with a number of industrial companies about their practices of using simulations for decision support.Based upon the analysis, a new, interactive DSS that can fulfill the industrial requirements, is proposed.The design of the cloud-based system architecture of the DSS is then described. To show the functionality and potential of the proposed DSS, an application study has been performed for the optimal design of a hypothetical but realistic flexible production cell. How important knowledge with respect to different preferences of the decision maker can be generated as rules, using the new Flexible Pattern Mining algorithm provided in the DSS, will be revealed by the results of this application study.

The health of a machine tool directly affects its ability to produce components with high precision. Therefore, monitoring and diagnosing early faults can enhance its reliability resulting in an improvement in manufacturing throughput and overall product quality. This paper concerns condition monitoring of the ballscrew drive, a machine tool component that transforms rotary motion of the drive shaft into linear motion of the work table along the guideways. The degradation of the ballscrew drive is often characterized by backlash, which results in imprecise linear motion and, therefore, affects the position of guideways during machining operations. Many physical characteristics of the ballscrew drive, such as required torque, viscous friction, and Coulomb friction, change with the degradation of the ballscrew during its lifetime. The paper proposes a condition monitoring methodology consisting of four main steps: data collection, data preprocessing and feature engineering, model building, and anomaly detection. The machine tool drive system is operated under no-load condition at regular intervals to capture health data using Siemens Analyze MyCondition instrumentation. Subsequently, the data is preprocessed and features are extracted from raw signals using the wavelet transform approach. The unsupervised machine learning technique, principal component analysis, is used to reduce the dimensionality of the dataset and find feature combinations that capture most of the variation in the data. Next, Hotelling’s T2 statistic is computed for each sample on a rolling basis, and anomalous behavior is detected based consistent deviations beyond the moving median of Hotelling’s T2 statistic. The proposed methodology is applied on condition monitoring data from a Swedish automotive manufacturer and the health assessments are validated against backlash measurements obtained from a different conditional monitoring test. This shows that the health status of a ballscrew can be derived directly from its physical characteristics.

Digitalization through Industry 4.0 technologies is one of the essential steps for the complete collaboration, communication, and integration of heterogeneous resources in a manufacturing organization towards improving manufacturing performance. One of the ways is to measure the effective utilization of critical resources, also known as bottlenecks. Finding such critical resources in a manufacturing system has been a significant focus of manufacturing research for several decades. However, finding a bottleneck in a complex manufacturing system is difficult due to the interdependencies and interactions of many resources. In this work, a digital twin framework is developed to detect, diagnose, and improve bottleneck resources using utilization-based bottleneck analysis, process mining, and diagnostic analytics. Unlike existing bottleneck detection methods, this novel approach is capable of directly utilizing enterprise data from multiple levels, namely production planning, process execution, and asset monitoring, to generate event-log which can be fed into a digital twin. This enables not only the detection and diagnosis of bottleneck resources, but also validation of various what-if improvement scenarios. The digital twin itself is generated through process mining techniques, which can extract the main process map from a complex system. The results show that the utilization can detect both sole and shifting bottlenecks in a complex manufacturing system. Diagnosing and managing bottleneck resources through the proposed approach yielded a minimum throughput improvement of 10% in a real factory setting. The concept of a custom digital twin for a specific context and goal opens many new possibilities for studying the strong interaction of multi-source data and decision-making in a manufacturing system. This methodology also has the potential to be exploited for multi-objective optimization of bottleneck resources.

Production systems are evolving rapidly, thanks to key Industry 4.0 technologies such as production simulation, digital twins, internet-of-things, artificial intelligence, and big data analytics. The combination of these technologies can be used to meet the long-term enterprise goals of profitability, sustainability, and stability by increasing the throughput and reducing production costs. Owing to digitization, manufacturing companies can now explore operational level data to track the performance of systems making processes more transparent and efficient. This untapped granular data can be leveraged to better understand the system and identify constraining activities or resources that determine the system’s throughput. In this paper, we propose a data-driven methodology that exploits the technique of data integration, process mining, and analytics based on factory physics to identify constrained resources, also known as bottlenecks. To test the proposed methodology, a case study was performed on an industrial scenario were a discrete event simulation model is built and validated to run future what-if analyses and optimization scenarios. The proposed methodology is easy to implement and can be generalized to any other organization that captures event data.

This paper presents a decision support approach to enable decision-makers to identify no-preference solutions in multi-objective optimization for factory layout planning. Using a set of trade-off solutions for a battery production assembly station, a decision support method is introduced to select three solutions that balance all conflicting objectives, namely, the solution closest to the ideal point, the solution furthest from the nadir point, and the one that is best performing along the ideal nadir vector. To further support decision-making, additional analyses of system performance and worker well-being metrics are integrated. This approach emphasizes balancing operational efficiency with human-centric design, aligning with human factors and ergonomics (HFE) principles and Industry 4.0–5.0. The findings demonstrate that objective decision support based on Pareto front analysis can effectively guide stakeholders in selecting optimal solutions that enhance both system performance and worker well-being. Future work could explore applying this framework with alternative multi-objective optimization algorithms.

Group decision making is traditionally a human-centered process, where communication, synchronization and agreement are driven by the stakeholders involved. In the area of multi-objective optimization (MOO), this becomes a challenge, because MOO usually produces a large amount of trade-off solutions that need to be analyzed and discussed by the stakeholders. Moreover, for transparent group decision making, it is important that each decision maker is able to trace the entire decision process – from associated data and models to problem formulation and solution generation, as well as to the preferences and analyses of other decision makers. A graph database is capable of capturing such diverse information in the form of a knowledge graph. It can be used to store and query all dependencies and hence can support complex decision-making tasks. Further advantages are the inherent suitability for visualization and the possibilities for pattern matching, graph analytics and, if semantically enriched, to infer new connections in the graph. In this paper, we show how such a knowledge graph can be used to support more transparent and traceable decision-making activities, particularly when multiple stakeholders with differing preferences or perspectives are involved. 

This paper conceptually introduces VF-KDO (Virtual Factories with Knowledge-Driven Optimization, a research profile of the University of Skovde, Sweden, which is underway from 2018-2026. The goal of this research profile is to deliver radical innovations in manufacturing research essential to the design and operation of next-generation manufacturing systems. A unique concept proposed in VF-KDO is: knowledge extracted for decision support is achieved through systematically exploring, e.g., using advanced, interactive data analytics techniques on optimal solutions generated via many-objective optimizations on virtual factory models. As the word 'driven' means 'motivated' or 'manipulated', so does KDO have some two-fold meanings: (1) optimizations that aim at generating knowledge, not only mathematically optimal solutions; (2) knowledge-controlled optimizations, instead of some blind/black-box processes. It is this concept of KDO, combining with modular, virtual factory models at different levels, which distinguishes VF-KDO from other related research efforts found internationally and in Sweden. The cutting-edge research topics involved in the research profile and their synergy with the digitalization efforts of the 7 partner companies, in form of the development of an intelligent decision support system, can be used to improve the competiveness of the Swedish manufacturing industry by supporting their holistic, optimal and sustainable decision making. 

Assembly processes have the most influencing and long-term impact on the production volume and cost in the aerospace industry. One of the most crucial factors in aircraft assembly lines design during the conceptual design phase is ramp-up planning that synchronizes the production rates at the globally dispersed facilities. Inspired by a pilot study performed with an aerospace company, this paper introduces a hybrid simulation-optimization approach for addressing an assembly production chain ramp-up problem that takes into account: (1) the interdependencies of the ramp-up profiles between final assembly lines and its upstream lines; (2) workforce planning with various learning curves; (3) inter-plant buffer and lead-time optimization, in the problem formulation. The approach supports the optimization of the ramp-up profile that minimizes the times the aircraft assemblies stay in the buffers and simultaneously attains zero backlog. It also generates the required simulation-optimization data for supporting the decision-making activities in the industrialization projects. 

This paper presents an innovative approach for the design and analysis of production systems using multi-objective optimization and data mining. The innovation lies on how these two methods using different computational intelligence algorithms can be synergistically integrated and used interactively by production systems designers to support their design decisions. Unlike ordinary optimization approaches for production systems design which several design objectives are linearly combined into a single mathematical function, multi-objective optimization that can generate multiple design alternatives and sort their performances into an efficient frontier can enable the designer to have a more complete picture about how the design decision variables, like number of machines and buffers, can affect the overall performances of the system. Such kind of knowledge that can be gained by plotting the efficient frontier cannot be sought by single-objective based optimizations. Additionally, because of the multiple optimal design alternatives generated, they constitute a dataset that can be fed into some data mining algorithms for extracting the knowledge about the relationships among the design variables and the objectives. This paper addresses the specific challenges posed by the design of discrete production systems for this integrated optimization and data mining approach and then outline a new interactive data mining algorithm developed to meet these challenges, illustrated with a real-world production line design example.

In recent years, interactive evolutionary multiobjective optimization methods have been getting more and more attention. In these methods, a decision maker, who is a domain expert, is iteratively involved in the solution process and guides the solution process toward her/his desired region with preference information. However, there have not been many studies regarding the performance evaluation of interactive evolutionary methods. On the other hand, indicators have been developed for a priori methods, where the DM provides preference information before optimization. In the literature, some studies treat interactive evolutionary methods as a series of a priori steps when assessing and comparing them. In such settings, indicators designed for a priori methods can be utilized. In this paper, we propose a novel performance indicator for interactive evolutionary multiobjective optimization methods and show how it can assess the performance of these interactive methods as a whole process and not as a series of separate steps. In addition, we demonstrate the shortcomings of using indicators designed for a priori methods for comparing interactive evolutionary methods. IEEE

Interactive methods support decision makers in finding the most preferred solution in multiobjective optimization problems. They iteratively incorporate the decision maker's preference information to find the best balance among conflicting objectives. Several interactive methods have been developed in the literature. However, choosing the most suitable interactive method for a given problem can prove challenging and appropriate indicators are needed to compare interactive methods. Some indicators exist for a priori methods, where preferences are provided at the beginning of the solution process. We present some numerical experiments that illustrate why these indicators are not suitable for interactive methods. As the main contribution of this paper, we propose a set of desirable properties of indicators for assessing interactive methods as the first step of filling a gap in the literature. We discuss each property in detail and provide simple examples to illustrate their behavior. 

Although an entire frontier of Pareto-optimal solutions exists for multi-objective optimization problems, in practice, decision makers are often only interested in a small subset of these solutions, called the region of interest. Specialized optimizers, such as reference-point based evolutionary algorithms, exist that can focus the search to only find solutions inside this region of interest. These algorithms typically only modify the selection mechanism of regular multi-objective optimizers to preferentially select solutions that conform to the reference point. However, a more effective search may be performed by additionally modifying the variation mechanism of the optimizers, namely the crossover and the mutation operators, to preferentially generate solutions conforming to the reference point. In this paper, we propose a modular mutation operator that uses a recent knowledge discovery technique to first find decision rules unique to the preferred solutions in each generation. These rules are then used to build an empirical distribution in the decision space that can be sampled to generate new mutated solutions which are more likely to be closer to the preferred solutions. The operator is modular in the sense that it can be used with any existing reference-point based evolutionary algorithm by simply replacing the mutation operator. We incorporate the proposed knowledge-driven mutation operator into three such algorithms, and through benchmark test problems up to 10 objectives, demonstrate that their performance improves significantly in the majority of cases according to two different performance indicators. 

The use of evolutionary algorithms for solving multi-objective optimization problems leaves the decision makers with a set of Pareto-optimal solutions to be considered in their decision making. Multi-objective optimization problems offer two spaces for a decision maker to analyze, the decision space and the objective space. In the literature, most of the focus has been on analyzing the objective space, however, in this paper, two procedures are presented for analyzing the decision space by identifying the variables that predominantly influence the structure of the objective space. Both procedures employ a recently proposed rule mining approach, which is used to find significant rules in terms of the variables. The rules are then combined and an influence score is calculated. The method is demonstrated on four problems, two scalable test problems (DTLZ2 and WFG2) with cases of three, five and seven objectives, one engineering design problem and one simulation-based optimization problem. The experiments show that the proposed approach is able to identify influential variables in most problem cases. 

In many practical applications, the end-goal of multi-objective optimization is to select an implementable solution that is close to the Pareto-optimal front while satisfying the decision maker’s preferences. The decision making process is challenging since it involves the manual consideration of all solutions. The field of multi-criteria decision making offers many methods that help the decision maker in this process. However, most methods only focus on analyzing the solutions’ objective values. A more informed decision generally requires the additional knowledge of how different preferences affect the variable values. One difficulty in realizing this is that while the preferences are often expressed in the objective space, the knowledge required to implement a preferred solution exists in the decision space. In this paper, we propose a decision support system that allows interactive knowledge discovery and knowledge visualization to support practitioners by simultaneously considering preferences in the objective space and their impact in the decision space. The knowledge discovery step can use either of two recently proposed data mining techniques for extracting decision rules that conform to given preferences, while the extracted knowledge is visualized via a novel graph-based approach that allows the discovery of important variables, their values and their interactions with other variables. The result is an intuitive and interactive decision support system that aids the entire decision making process — from solution visualization to knowledge visualization. We demonstrate the usefulness of this system on benchmark optimization problems up to 10 objectives and real-world problems with up to six objectives.

Practical multi-criterion decision making not only involves the articulation of preferences in the objective space, but also a consideration of how the variables impact these preferences. Trend mining is a recently proposed visualization technique that offers the decision maker a quick overview of the variables' effect on the structure of the objective space and easily discover interesting variable trends. The original trend mining approach relies on a set of predefined reference directions along which an interestingness score is measured for each variable. In this paper, we relax this requirement by automating the approach to find optimal reference directions that maximize the interestingness for each variable. Additional extensions include the use of an Achievement Scalarizing Function (ASF) for ranking solutions along a given reference direction, and an updated interestingness score formulation for more appropriately handling discrete variables. We demonstrate the working of the extended approach on DTLZ2 and WFG2 benchmarks for up to five objectives and on a biobjective engineering design problem. The results show that the ability of the proposed approach to detect variable trends in high dimensional objective spaces is heavily dependent on the quality of the solutions used. 

Practitioners of multi-objective optimization currently lack open tools that provide decision support through knowledge discovery. There exist many software platforms for multi-objective optimization, but they often fall short of implementing methods for rigorous post-optimality analysis and knowledge discovery from the generated solutions. This paper presents Mimer, a multi-criteria decision support tool for solution exploration, preference elicitation, knowledge discovery, and knowledge visualization. Mimer is openly available as a web-based tool and uses state-of-the-art web-technologies based on WebAssembly to perform heavy computations on the client-side. Its features include multiple linked visualizations and input methods that enable the decision maker to interact with the solutions, knowledge discovery through interactive data mining and graph-based knowledge visualization. It also includes a complete Python programming interface for advanced data manipulation tasks that may be too specific for the graphical interface. Mimer is evaluated through a user study in which the participants are asked to perform representative tasks simulating practical analysis and decision making. The participants also complete a questionnaire about their experience and the features available in Mimer. The survey indicates that participants find Mimer useful for decision support. The participants also offered suggestions for enhancing some features and implementing new features to extend the capabilities of the tool.

Current market requirements force manufacturing companies to face production changes more often than ever before. Reconfigurable manufacturing systems (RMS) are considered a key enabler in today's manufacturing industry to cope with such dynamic and volatile markets. The literature confirms that the use of simulation-based multi-objective optimization offers a promising approach that leads to improvements in RMS. However, due to the dynamic behavior of real-world RMS, applying conventional optimization approaches can be very time-consuming, specifically when there is no general knowledge about the quality of solutions. Meanwhile, Pareto-optimal solutions may share some common design principles that can be discovered with data mining and machine learning methods and exploited by the optimization. In this study, the authors investigate a novel knowledge-driven optimization (KDO) approach to speed up the convergence in RMS applications. This approach generates generalized knowledge from previous scenarios, which is then applied to improve the efficiency of the optimization of new scenarios. This study applied the proposed approach to a multi-part flow line RMS that considers scalable capacities while addressing the tasks assignment to workstations and the buffer allocation problems. The results demonstrate how a KDO approach leads to convergence rate improvements in a real-world RMS case.

In order to pursue the dream combination of human flexibility and robot automation, human robot collaboration (HRC) is increasingly being investigated through academic research and industrial scenarios. HRC involves several challenges ranging from safety and comfort of the human to process efficiency and cost of robot operation. Achieving the right balance between these aspects is critical to implementing a safe, profitable and sustainable HRC environment. In this paper,we propose the use of simulation-based optimization (SBO) for assembly task allocation and scheduling for a HRC working cell in which an industrial robot assists a human worker. The list of product assembly operations are classified according to the capability of human worker and robot, and the sequencing constraints on them are the initial inputs of the method. The operators’ ergonomic load scores and cycletime of the assembly process are achieved by simulation. The optimized solutions are sorted to find the trade-offs between ergonomics and cycle time. We demonstratethe feasibility of the proposed approach through an industrial case study.

Genetic algorithms have been used as an optimization tool using evolutionary strategies. Genetic algorithms cover three basic steps for population refinement selection, cross-over and mutation. In normal Real-coded genetic algorithm (RGA), the population of real variables generated after population refinement operations, is used for the computation of the objective function. In this paper we have shown the effect made by mapping the refined population towards better solutions and thereby creating more biased search. The mapping used is non-uniform in nature and is the function of the position of the individual w.r.t. the best solution obtained so far in the algorithm, and hence the name Non-Uniform RGA or in short NRGA. Tests were performed on standard benchmark problems. The results were promising and should encourage further research in this dimension.