Approximation algorithms for the min-max regret identical parallel machine scheduling problem with outsourcing and uncertain processing time

Abstract

We consider the robust (min-max regret) version of identical parallel machine scheduling problem, in which jobs may be outsourced to balance total cost against production efficiency. The total cost is measured in terms of the total completion time of jobs processed in-house and the cost of outsourcing the rest. Processing times of in-house jobs are uncertain and they are described as two types of scenarios: discrete and interval. The objective is to obtain a robust (min-max regret) decision that minimises the absolute deviation of total cost from the optimal solution under the worst-case scenario. We first prove the worst-case scenario for any feasible solution. For the interval scenario, we further prove that the maximum regret value can be obtained in polynomial time for any feasible schedule. We also prove that for any discrete scenario, the minimum total cost can be obtained in polynomial time. Since the problem with the interval scenario is strongly NP-hard, we then transform the problem into an equivalent robust single machine scheduling problem. Finally, we develop 2-approximation algorithms for the problem with discrete and interval scenarios, respectively. These results are helpful for bridging the scheduling theory and practice in identical parallel machining environments with outsourcing and uncertain processing times.