Scientia Iranica
Volume:29 Issue: 2, Mar & Apr 2022

Companies must determine the replacement time of machine parts correctly since it affects their production costs and efficiencies. For this, it is aimed to determine the most appropriate replacement time to minimize cost per unit. In this study, it is proposed to develop a hybrid Artificial Neural Network (ANN)-Genetic Algorithm (GA) model to predict replacement time without using a cost model. At first, a replacement cost model is developed to calculate replacement times to use in training the neural network. Nevertheless, the cost model needs complex mathematical calculations. GA is used instead of the cost model to determine replacement time, and thus, to achieve fast learning for the neural network. The hybrid ANN-GA model was applied to predict replacement time of bladder in tire manufacturing. Furthermore, ANN and GA models, which were developed to increase the prediction accuracy of the hybrid model, were used. The hybrid ANN-GA model showed better performance with higher R^2 (0.943) and lower RMSE (9.124) and MAPE (2.528) values than the other ANN and GA models. The values indicate that the hybrid model is in good agreement with the cost model. Thus, it is recommended that the hybrid model is used instead of the cost model.

Different maintenance policies, including preventive maintenance and predictive maintenance, are introduced to enhance the execution of systems. Maintenance professional experts have faced numerous challenges with distinguishing the proper maintenance policy, among which causes of failure, accessibility, and the capability of maintenance should be regarded seriously. Moreover, most organizations do not have a deliberate and compelling model for evaluating maintenance policies under uncertainty to deal with real-world conditions. The aim of this paper is to introduce a new interval-valued fuzzy (IVF) decision model for the selection of maintenance policy based on order inclination with comparability to ideal solutions by Monte Carlo simulation. This paper introduces novel separation measures and a new IVF-distinguish index via possibilistic statistical concepts (PSCs) which can assist maintenance decision makers to rank maintenance policy candidates. Also, resilience engineering (RE) factors are considered along with conventional evaluation criteria. Finally, the steps of the proposed IVF model-based PSCs are applied to survey a real case in manufacturing industry. Results of the presented model are compared with the recent literature and could help maintenance personnel in identifying the best policy systematically.

This research paper builds a manufacturer-retailer integrated inventory model to compute jointly the optimal values for the order quantity, the lead time, the reorder point and the number of shipments taking into consideration the effect of learning-forgetting phenomenon on the setup cost. The fabrication process of manufacturer is not perfect and certain level of product quality can attain with an additional cost. Service level constraint is incorporated into the inventory model to evade the backorder which gives negative impact to company reputation. The lead time is reduced with the help of crashing cost. The proposed inventory model is illustrated with the help of an example. From this example, it is detected that centralized decision is better than decentralized one. It is also observed from the analysis that players have to compromise with their profit if they decide to increment the service level and quality of the product. Due to the effect of learning-forgetting on the setup cost, profit of the centralized system increases.

A bi-objective non-linear optimization model with the goal of maximizing the profit of inventory and minimizing the carbon emissions of transportation, simultaneously, is developed. Another contribution of this work is to propose three capable metaheuristics to solve it optimality in large-scale samples. In this regard, the Non-dominated Sorting Genetic Algorithm (NSGA-II) as a well-known method as well as Multi-Objective of Keshtel Algorithm (MOKA) and Multi-Objective of Red Deer Algorithm (MORDA) are firstly applied in this research area. The results of metaheuristics are checked by the ε-constraint method in a set of small-scale samples as compared with the results of literature. Finally, the outputs confirm that the allowed shortage situation along with the lack of cost reduction shows a greater amount of shipping and orders. As such, the performance of MORDA is approved in comparison with MOKA and NSGA-II through different criteria.

The product pricing decision is one of the important factors in the profitability of organizations, which has a key role in their survival. Moreover, the pricing is determined based on the demand and location of applicants. Therefore, the location of facilities and services influence the pricing. Also, the location problem is a critical issue in the survival of the organization. Furthermore, the hub location problem is a type of location problems, which has many applications and saves time and money. On the other hand, location is impossible without considering transportation and transportation has many negative effects on environmental, such as greenhouse gas emissions, air pollution, noise, etc. Considering this reason, it is important to consider the environmental costs to reduce the adverse effects. In this paper, we consider integrated pricing and hub location problem with environmental costs in the competitive market that customer choice is calculated according to the logit model (LM). We use a genetic algorithm (GA) to solve and observe the environmental cost, entrant profit, incumbent income, the impact of customer sensitivity and discount between hubs on the entrant profit. As a final point, the computational experiments demonstrate that the suggested GA is both efficient and effective.

In many countries, a rail network consists of a series of single lines with sidings where inter-train interactions (meeting, passing) occur. An effort has been made in this paper to study two of thesenetwork-related issues: 1) scheduling freight trains in a single-line corridor while ensuring the interactions to happen safely and 2) allocating freight to the scheduled trains considering the freight due/release date and train’s weight/capacity. To better illustrate the real-world freighttrains’ scheduling problems, both the scheduling and allocation problems have been addressed. Minimizing the trains’ traveling time, allocating maximum freight to the scheduled ones, and minimizing the total freight tardiness at the related destination are the objective functions ofthis study. Both problems and their solutions have been addressed separately using integer linear programming models, but an integrated novel heuristic algorithm has been proposed to solve them. The computational results demonstrated through a generated data set show both the modelvalidation and the efficiency of the heuristic algorithm. This heuristic algorithm has been so designed to incorporate the practical operational railway rules with modest modification and although its outputs slightly differ from the exact solutions, it can solve both models simultaneouslyin large scale problems.

Higher dimensional accuracy along with better surface finish of various advanced engineering materials has turned out to be the prime desideratum for the present day manufacturing industries. To achieve this, non-traditional machining (NTM) processes have become quite popular because of their ability to produce intricate shape geometries on diverse difficult-to-machine materials. To allow these processes to operate at their fullest capability, it is often recommended to set their different input parameters at the optimal levels. Thus, in this paper, a new technique combining grey correlation method and evaluation based on distance from average solution is applied for simultaneous optimization of three NTM processes, i.e. photochemical machining process, laser-assisted jet electrochemical machining process and abrasive water jet drilling process. The derived optimal parametric combinations outperform those as identified by the other popular multi-objective optimization techniques with respect to the considered response values. The results of analysis of variance also identify the most influencing parameters for the said NTM processes. Finally, the developed surface plots would help the process engineers in investigating the effects of different NTM process parameters on the corresponding grey appraisal scores.

In order to monitor mean and variability of a process, the Gini controlcharts based on the skew-normally distributed random samples is proposed. Throughcomparing the false alarm rates of current scheme with those of existing mean anddispersion control charts, it is found out that the design structure of Gini chart canimprove over other classic schemes based on assumption of skew-normal distributionfor the data. Moreover, the superiority of the Gini chart is studied by comparingthe discriminatory power curves of the skew-normal distribution with some existingcontrol charts. Simulated studies and a real data example illustrate the usefulness ofthe proposed approach.

The q-rung orthopair fuzzy set (q-ROFS) as a generalization of fuzzy set (FS), is characterized by membership and non-membership, and the sum of their q-powers is restricted to [0,1]. In this manuscript, a new Complex q-rung orthoapir fuzzy set (Cq-ROFS) is proposed by combining q-ROFS and complex fuzzy set (CFS). Cq-ROFS is a better way to process uncertain and imprecise information in decision making, which is characterized by complex-valued membership and complex-valued non-membership. First, some fundamental operational laws, score function and accuracy function, and comparison method are proposed. Further, because the vector similarity measures (VSMs) play a key role in statistics, physics and engineering sciences, some VSMs called Jaccard similarity measures (JSMs), dice similarity measures (DSMs) and cosine similarity measures (CSMs) for Cq-ROFSs and interval-valued complex q-rung orthopair fuzzy sets (IVCq-ROFSs) are investigated. Moreover, the hybrid vector similarity measures (HVSMs) called variation co-efficient similarity measures (VCSMs) for Cq-ROFSs and IVCq-ROFSs are also proposed and their properties are discussed. Finally, in order to demonstrate the feasibility of the investigated HVSMs, the existing similarity measures about complex Pythagorean fuzzy sets (CPFSs) and complex intuitionistic fuzzy sets (CIFSs) are compared with the proposed methods by numerical examples of medical diagnosis and pattern recognition.

Data envelopment analysis (DEA) technique is widely applied for performance assessment of decision making units (DMUs). The revenue efficiency (RE) evaluation is one of the controversial subject matters that can be performed through DEA context. The amount of productions and its prices are crucial factors in the RE. The classical DEA models consider the prices to be fixed and known which are not the case in real world. Also, the classical DEA models considers linear pricing in revenue assessment. However, most of real world problems deal with nonlinear prices. This paper evaluates the RE given the piecewise linear theory in non-competitive situations. In doing so, a stepwise pricing function is introduced which lets the prices to be changed in relation to the amount of the production. As an innovative idea, a more accurate mathematical modeling for the RE is proposed. We define a dynamic weights’ function in maximum revenue optimization model which no longer considers fixed prices. A case study validates our proposed model.

The aim of this study is to investigate causality patterns of safety-related incidents in the construction industry. Although there are many studies to find cause-and-effect relationships in the accident database, retrieving useful knowledge from the last database and taking additional variables into account are needed. Therefore, in the present study, the association rule method was utilized to investigate a large number of historical accident data in Iran's construction industry in the duration of 2014-2017. Based on association rules results, the combination of worker's individual and behavioral factors and supervisory conditions are more related to serious accidents. These results can provide practical insights for construction managers who need to be more concerned about the negative effects of the combination of some factors on serious construction accidents.