This special issue of Engineering Optimization (EO) originates from the ‘Asia Pacific Industrial Engineering and Management Systems (APIEMS) 2016 Conference’ that was held in Taipei, Taiwan, between 7 and 10 December 2016. APIEMS 2016 was co-hosted by the Department of Industrial Management, Taiwan Tech, and Department of Industrial Engineering and Management, Taipei Tech, Taiwan.
The idea of editing this special issue was initiated with the planning of the APIEMS 2016. The primary objective is to provide a great opportunity for the participants to share their knowledge as well as the problems related to supply chain management. Eight articles made it through two or more rigorous peer-review cycles according to the standards of the EO journal.
Articles in the special issue
In the first article, ‘Consultation sequencing of a hospital with multiple service points using genetic programming’, Morikawa, Takahashi and Nagasawa propose a genetic programming (GP) approach to the consultation sequencing problem with multiple service points. The objective is to minimize the sum of three weighted average waiting times of scheduled, walk-in and second round consultation patients. They decide the dispatching rules for selecting the next patient to enter the consultation room. Simulation results showed that dispatching rules produced by GP to prioritize different queues and display the consultation sequence could reduce the average waiting time for patients.
In the next article, ‘Prepositioning emergency supplies under uncertainty: a parametric optimization method’, Bai, Gao and Liu formulate a robust fuzzy value-at-risk (VaR) optimization model for the prepositioning emergency supplies problem. The objective function and constraints were turned into equivalent parametric forms through chance-constrained programming. They developed a parameter-based domain decomposition method to divide the problem into six mixed-integer parametric sub-models that were solved by LINGO. The computational results showed the credibility and superiority of this method over compared studies.
In the third article, ‘Comparison of particle swarm optimization and differential evolution for aggregators’ profit maximization in demand response system’, Wisittipanit and Wisittipanich discuss the smart grid demand response (DR) problem with three stakeholders: a utility operator, a set of aggregators and end-customers. The problem is formulated as a multilevel optimization problem: the utility operator aims to minimize its operational cost, the aggregator to maximize its net profit and the end-customer to maximize its net payoff. Two metaheuristics, particle swarm optimization (PSO) and differential evolution (DE), were proposed and compared. DE provided better results than PSO in the numerical example.
In ‘An iterated local search algorithm for the team orienteering problem with variable profits’, Gunawan, Ng, Kendall and Lai formulate a mathematical model for the team orienteering problem with variable profits (TOPVP). The objective is to maximize the collected score, which depends on the duration of stay on the visited vertex. An iterated local search (ILS) with record-to-record travel acceptance criterion was developed to solve the TOPVP and compared with the optimization software AIMMS for modified benchmark instances. The ILS could obtain optimality in several small-sized instances and good solutions in large-sized instances with shorter computational time.
In the fifth article, ‘An economic order quantity model with nonlinear holding cost, partial backlogging and ramp-type demand’, San-José, Sicilia, González-de-la-Rosa and Febles-Acosta develop an inventory model that combines ramp-type demand and power demand patterns with a variable inventory cycle and nonlinear holding cost depending on holding time. The objective is to maximize the total profit per unit time. They proposed an exact solution procedure to find the optimal economic lot size and inventory cycle that allow fractional backlogged demand. The results of three examples with different input parameters illustrated the theoretical properties.
In the article ‘Simulated annealing heuristic for general share-a-ride problem’, Yu, Purwanti, Perwira Redi, Lu, Syprayogi and Jewpanya extend the share-a-ride problem (SARP) to allow ride sharing between passengers, called the general share-a-ride problem (G-SARP). The G-SARP was first formulated as a mixed-integer programming model to maximize total profit. Because the G-SARP is NP hard, this study developed a simulated annealing algorithm to solve the problem. Through two benchmark data sets, they compared the results with CPLEX and the dial-a-ride solution as the lower bound. The results showed that the proposed SA could find optimal solutions in small-sized instances and better objective values than the original SARP.
In the seventh article, ‘Early stage response problem for post-disaster incidents’, Kim, Shin, Lee, and Moon consider an early emergency response problem with time-dependent risks to maximize prevented risks during the operation for a limited number of responders. They formulated the problem with a mixed-integer programming model. Two greedy algorithms were developed to solve the problem. To evaluate the model and algorithms, experimental data from a high-rise building were tested. The results from the proposed greedy algorithms were compared with optimization software and sequential algorithms. The proposed greedy algorithm could find near-optimal solutions with shorter computational time.
In the last article, ‘Network reliability maximization for stochastic-flow network subject to correlated failures using genetic algorithm and tabu search’, Yeh, Lin and Yang determine the optimal resource assignment with maximal network reliability for stochastic-flow networks (SFNs). They took into account the correlated failure in the network reliability optimization. To solve such an NP-hard problem, a hybrid algorithm integrating genetic algorithm (GA) and tabu search (TS), called the hybrid GA-TS algorithm (HGTA), was proposed. Four numerical examples were tested and compared with other soft computing algorithms. The results showed that HGTA can obtain high-quality solutions with efficient computational times.
Acknowledgements
We sincerely thank all of the authors for presenting and submitting their original research articles for consideration in this special issue, and all of the anonymous referees for their valuable time and high-quality reviews. Without their support and cooperation, this special issue could not have been published in such a timely manner. We also appreciate the support and guidance of the Engineering Optimization Editorial Office.