Journal of Advances in Industrial Engineering
Volume:53 Issue: 4, Autumn 2019

Most humanitarian relief items' investigations try to satisfy demands in disaster areas in an appropriate time and reduce the rate of causality. Time is an essential element in humanitarian relief items; the quietest response time, the more rescued people. Reducing response time with high reliability is the main objective of this research. In our investigation, monitoring the route’s situation after occurrence disaster with drones and motorcycles is planned for collecting information about routes and demand points in the first stage. The collected information is analyzed by the disaster management to determine the probability of each scenario. By evaluating collected data, the route repair groups are sent to increase the route’s reliability. In the final step, the relief items operation allocates the relief items to demand points. All in all, this research tries to present a practical model and real situation to survive more people after occurrence disaster. An exact solver solves the evolutionary model in small and medium scales; the developed model in big scale is solved by Grasshopper Optimization Algorithm (GOA), and then results are evaluated. The evaluation results indicate the positive effect of valid initial information on the humanitarian supply chain’s performance.

This paper investigates steel-making continuous casting (SCC) scheduling problem. SCC is a high temperature and large-scale logistics machining process with batch production at the last stage that was identified as the key process of modern iron and steel enterprises. This paper presents a mathematical model for scheduling SCC process. The model is developed as a Mixed Zero- One Linear programming (MZOLP) based on actual production situations of SCC. The objective is to schedule a set of charges (jobs) to minimize the earliness and tardiness penalty costs as well as the charge waiting time cost. The solution methodology is developed based on a branch-and-bound algorithm. A heuristic method is presented at the beginning of the search in order to compute an initial upper bound. A lower bound and an upper bound are developed and a method for reducing branches is established based on the batch production in the continuous casting (CC) stage. Moreover, branching schemes are proposed. The branch- and- bound algorithm incorporating the initial upper bound, the lower and upper bound, the method for reducing branches, and branching schemes is tested on a set of instances. The analysis shows the efficiency of the proposed features for the algorithm.

Natural and technological disasters threaten human life all around the world significantly and impose many damages and losses on them. The current study introduces a multi-objective three-stage location-routing problem in designing an efficient and timely distribution plan in the response phase of a possible earthquake. This problem considers uncertainty in parameters such as demands, access to routes, time and cost of travels, and the number of available vehicles. Accordingly, a three-stage stochastic programming approach is applied to deal with the uncertainties. The objective functions of the proposed problem include minimizing the unsatisfied demands, minimizing the arriving times, and minimizing the relief operations costs. A modified algorithm of the improved version of the augmented ε-constraint method, which finds Pareto-optimal solutions in less computational time, is presented to solve the proposed multi-objective mixed-integer linear programming model. To validate the model and evaluate the performance of the methods several test problems are generated and solved by them. The computational results show the satisfactory performance of the proposed methods and effectiveness of the proposed model for delivery of relief commodities in the affected areas.

Accurate demand forecasting plays an important role in meeting customers’ expectations and satisfaction that strengthen the enterprise's competitive position. In this research, time series and artificial neural networks methods compete to provide more precise demand estimation while having a large variety of products. After obtaining the initial results, suggestions have been implemented to improve forecasting accuracy. As a direct result of that, the average of mean absolute percentage error (MAPE) of all products' demand forecast reduces significantly. To improve the quality of historical records, association rules and substitution ratio have been applied . This method plays a significant role to detect the existing pattern in historical data and MAPE reduction. The satisfactory and applicable results provide the company with more accurate forecast. Moreover, the issue of precepting confusing historical data which caused unforecastable trends has been solved. The R language and “neuralnet”, “nnfor”, “forecast”, and “arules” packages have been applied in programming.

In this study, a robust optimization model is introduced, we propose a location-routing problem with simultaneous pickup and delivery under a hard time window that has a heterogeneous and limited depot and vehicle capacities and multi-variety of products and uncertain traveling time that considering all of these constraints together make the problem closer to real practical world’s problems, that not been studied in previous papers. For this purpose, a mixed-integer linear programming (MILP) model is proposed for locating depots and scheduling vehicle routing with multiple depots. Then, the robust counterpart of the proposed MILP model is proposed. The results show that the GA performs much better than the exact algorithm concerning time. GAMS software fails to solve the large-size problem, and the time to find a solution grows exponentially with increasing the size of the problem. However, the GA quite efficient for problems of large sizes, and can nearly find the optimal solution in a much shorter amount of time. Also, results in the Robust model show that increasing the confidence level has led to an increase in the value of the objective function of the robust counterpart model, this increase does not exhibit linear behavior. At 80% confidence level, the minimum changes in the objective function are observed, if we want to obtain a 90% confidence level, it requires more cost, but increasing the confidence level from 70% to 80% does not need more cost, so an 80% confidence level can be considered as an ideal solution for decision-makers.

Due to facing an acute shortage of beds in hospitals, the danger of getting involved in hospital infections and high-cost hospitals care, the Home Health Care industry has encountered high demands in recent years. Different stakeholders with various interests are involved in home health care that makes the process of planning and scheduling of nurses, who offered services, challenging. This paper, therefore, focuses on scheduling and routing nurses traveled to the patient’s home by considering the main features of the problem such as Continuity of Care and temporal dependencies. A new formulation for adjusting the time distance between two consecutive jobs performed by a nurse is presented. A feasible solution has to consider nurse and patient’s preferences, time windows for jobs, nurse’s qualification, and waiting time. A genetic algorithm is proposed to solve the problem. The computational results show the efficiency of the proposed algorithm, especially for large-size instances. Finally, the effect of the nurse’s dispatching policy on the objective function, waiting, and traveling times is examined.