Journal of the Operational Research Society

Gou, Xunjie; Xu, Zeshui; Liao, Huchang; Herrera, Francisco

doi: 10.1080/01605682.2020.1806741pmid: N/A

Abstract Double hierarchy hesitant fuzzy linguistic term set (DHHFLTS) can be used to express complex linguistic information by combining two hierarchy linguistic term sets with 2-tuple linguistic structures. In decision-making processes, experts’ assessment information may often be represented by some possible double hierarchy hesitant fuzzy linguistic elements (DHHFLEs) or some DHHFLEs with probability information, and we cannot ignore these probabilities when they are directly provided or aggregated by the experts’ assessments. As we are aware that representing probability information is a new improvement and challenge for DHHFLTSs, this paper defines a novel and more general concept named probabilistic double hierarchy linguistic term set (PDHLTS). Then, to propose some more reasonable operations and a distance measure of PDHLTSs, we develop an adjustment method to ensure that two PDHLTSs have same probability distribution. Additionally, this paper develops an extended probabilistic double hierarchy linguistic VIKOR method by improving the traditional VIKOR method. Moreover, the advantages and practicality of the proposed method are demonstrated by applying it to solve a practical multiple criteria decision-making problem involving smart healthcare. Finally, we make some comparative analyses, as well as discussing possible directions for future studies.

Aydas, Osman T.; Ross, Anthony D.; Scanlon, Matthew C.; Aydas, Buket

doi: 10.1080/01605682.2020.1806742pmid: N/A

Abstract This work examines medium-term integrated nurse staffing policy options for hospital Intensive Care Units (ICU) Our aim is to reduce nurse staffing costs while balancing the under/ over-staffing risks. Medium-term nurse schedules are highly uncertain as they are generated long before actual patient demand is realised. Optimisation models presented in this study allow us to examine fixed versus dynamic nurse staffing policy options for the medical units. In the dynamic nurse staffing, we utilise historical patient data to fit estimates of non-stationary patient demand. We compare the performance of both policy options with the optimal staffing scheme reached by the actual patient data. We generate feasible schedules for nurse sub-groups to avoid complete enumeration. We evaluate the performance of models with the pediatric ICU of a large urban children’s hospital. Experiments with the dynamic policy resulted in more than 3% higher average cost savings compared to the fixed staffing policies.

Masilionyte, Milda; McLean, Ian; Harding, Oliver; Doostmohammadi, Mahdi

doi: 10.1080/01605682.2020.1806743pmid: N/A

Abstract With the increasing number of patients suffering from knee osteoarthritis, the UK National Health Service is considering introducing a new treatment option that would focus on lifestyle changes. This study aims to develop a novel model that could serve as a tool to estimate the impact of such an intervention on treatment outcomes. In collaboration with the Forth Valley Royal Hospital, Larbert, United Kingdom, the model was formulated as a system dynamics simulation model and was built using Insight Maker, a web-based modelling tool. To the best of our knowledge, this paper is the first to employ system dynamics to tackle this problem. The simulations were run for several configurations to better understand the potential impact of advanced lifestyle treatment under various scenarios. The results for the most expected scenario suggest that introducing advanced lifestyle treatment would increase the average number of recovered patients by 4%, and reduce the average numbers of temporarily disabled, permanently disabled and deceased patients by 21%, 9% and 4%, respectively. The results also reveal that even with low advanced lifestyle treatment acceptance rates, the treatment outcomes could improve without any changes to current resources.

Liu, Xiaoxi; Kim, Byung Cho

doi: 10.1080/01605682.2020.1806744pmid: N/A

Abstract With fast-moving technological innovation, technology firms are confronted with the issue of how to dispose of the old technology-based product. Regarding the old technology-based product’s response strategy to a prevalent new technology-based product, firms usually choose from the three options of stopping, maintaining, and improving. We consider a multiproduct monopolist that manufactures traditional, tangible goods derived by old and new technologies, respectively, and build a two-period analytical model to form the response strategy for the old technology-based product. We also investigate the action sequence about whether to stop or improve the old product before or after network externality arises in the market, and determine the level of improvement effort exerted on the old product. The results show that the firm should maintain the old product if the new product is significantly superior to the old one. Furthermore, when the new product is superior to the old one, the firm has an incentive to stop or improve the old product based on the quality difference between old and new products and the marginal production cost of the old product. Finally, we prove that the improvement of old products can increase demands for both old and new products under specific conditions.

Sun, Xinyu; Geng, Xin-Na; Liu, Feng

doi: 10.1080/01605682.2020.1806746pmid: N/A

Abstract This article considers the flow shop problem of minimising the total weighted completion time in which the processing times of jobs are variable according to general position weighted learning effects. Two simple heuristics are proposed, and their worst-case error bounds are analysed. In addition, some complex heuristics (including simulated annealing algorithms) and a branch-and-bound algorithm are proposed as solutions to this problem. Finally, computational experiments are performed to examine the effectiveness and efficiency of the proposed algorithms.

Jeuken, Rick; Forbes, Michael; Kearney, Michael

doi: 10.1080/01605682.2020.1811167pmid: N/A

Abstract Coking coal is essential for the production of steel, and the quality of this coal significantly contributes to the quality of the produced steel. High quality coking coal has low ash content and a range of properties including volatile matter content and predicted coke strength. The coal is improved by processing after it has been mined. This processing varies and coal from multiple sources is blended. This paper introduces an original mixed integer programming model to maximise the profit of coal blending and processing. The model is computationally efficient and can be implemented at any coal mining and processing operation. The multi-period blending model incorporates stockpiling of raw material, and explicitly captures the geological variability of coal using chance constraints. A case study is evaluated and demonstrates that explicitly modelling geological variability can reduce the risk of breaching product specifications without any revenue loss. The improvement is achievable, without additional cost, by selecting the order that coal is fed into a processing plant.

Zhou, Zhongbao; Li, Kai; Liu, Qing; Tao, Zui; Lin, Ling

doi: 10.1080/01605682.2020.1811168pmid: N/A

Abstract The construction industry plays an important role in China's industrialization and urbanization process, which has become a major contributor to carbon emission in China. We innovatively use carbon footprint as an indicator of undesirable output in evaluating the construction industry since it captures both direct and indirect carbon emission. Then we conduct a comprehensive analysis of eco-efficiency of this industry under the framework of natural disposability and managerial disposability with a new radial DEA model on considering both desirable and undesirable outputs. We have several important findings. First, non-metallic mineral products, metal smelting and calendering, electricity and heat production, and supply are major drives from the perspective of the industrial chain. And the carbon footprint is mainly caused by capital formation from the final demand perspective. Second, most regions have a relatively lower eco-efficiency due to weak awareness of environmental protection, and there is a large emission reduction potential by eco-technology innovation. Finally, developed regions prefer to adopt the strategy of increasing investment in eco-technology innovation. The other regions generally adopted the strategy of reducing all inputs. The government should strengthen environment regulation and encourage the developing regions to increase investment in eco-technology innovation.

Fang, Lei; Zhao, Yuanyuan

doi: 10.1080/01605682.2020.1811169pmid: N/A

Abstract Taking the two-stage process that the outputs from the first stage are taken as the inputs for the second stage as example, internal resources conflict refers to the mismatch in amount of intermediate products between two efficient stages. Although the existing centralized models can eliminate this conflict by maximizing the efficiency of entire system. In many real-world situations, however, because of decentralized organizational structure of system and independent operational mechanism of stages, each decentralized stage wants to maximize its own efficiency. This leads to the internal resources conflict, which is called the risk of mismatch in this paper. Therefore, this paper proposes the risk sharing strategy to eliminate such conflict by sharing this risk between two decentralized stages, and the slacks-based measure mixed integer linear program (MILP) model is developed to ensure that the risk allocated to each decentralized stage is minimal. Finally, the proposed method is illustrated through two examples. The results show that the proposed method can not only identify the target intermediate products to solve the potential conflict, but also make the efficiency score of inefficient system higher than that of inefficient system measured by the centralized method of “free” link constraint.

Gao, Xuehong

doi: 10.1080/01605682.2020.1811790pmid: N/A

Abstract After a large-scale disaster occurred, medical service centers are needed to provide appropriate medical treatments for injured patients. To support those medical service centers, a warehouse should be set up to stock medical supplies that can be distributed to medical service centers in different quantities in multiple periods. However, the delivery of medical supplies depends on practical environmental conditions. In practice, the cost including the travel time and monetary cost from the warehouse to the medical service center presents a complex relationship. The existing methods do not consider this complex relationship. To overcome the pitiful of the existing methods, we propose a location-driven method. To implement the proposed location-driven method, three numerical algorithms are considered. Among them, the fixed-point-based heuristic algorithm is compared with the existing two algorithms to illustrate its superiority. Also, the bootstrap resampling technique is applied to obtain a reasonable region for improving the flexibility of locating the warehouse. To illustrate the effectiveness and efficiency of the proposed method, several simulation studies are considered. The computational results show the advantages of the proposed method in determining the warehouse location and identifying the reasonable region for locating the warehouse.

Martin, Mateus; Morabito, Reinaldo; Munari, Pedro

doi: 10.1080/01605682.2020.1813640pmid: N/A

Abstract In this article, we address the Constrained Two-dimensional Guillotine Cutting Problem (C2GCP) and the Constrained Three-dimensional Guillotine Cutting Problem (C3GCP). These problems consist of cutting a rectangular two-/three-dimensional object with orthogonal guillotine cuts to produce ordered rectangular two-/three-dimensional items seeking the most valuable subset of items cut. They often appear in manufacturing settings that cut objects to produce item types of low demand, such as in the cutting of flat glass in the glass industry, rocks in the granite and marble industries and steel blocks in the metallurgical industry. To model and solve these problems, we propose a novel top-down cutting approach that leads to effective mixed integer linear programming models for the C2GCP and the C3GCP. The insight of the proposed approach is to represent the cutting pattern as a binary tree, in which the root node is the object, and branches correspond to guillotine cuts. The results of computational experiments with a general-purpose optimization solver and using three sets of benchmark instances showed that the proposed models are competitive with state-of-the-art formulations of the C2GCP and the C3GCP in quality of solution and processing times, particularly when the number of items in an optimal solution is moderate.