Abstract
Optimizing the trade-off between crucial decisions has been a prominent issue to help decision-makers for synchronizing the production scheduling and distribution planning in supply chain management. In this article, a bi-objective integrated scheduling problem of production and distribution is addressed in a production environment with identical parallel machines. Besides, two objective functions are considered as measures for customer satisfaction and reduction of the manufacturer’s costs. The first objective is considered aiming at minimizing the total weighted tardiness and total operation time. The second objective is considered aiming at minimizing the total cost of the company’s reputational damage due to the number of tardy orders, total earliness penalty, and total batch delivery costs. First, a mathematical programming model is developed for the problem. Then, two well-known meta-heuristic algorithms are designed to spot near-optimal solutions since the problem is strongly NP-hard. A multi-objective particle swarm optimization (MOPSO) is designed using a mutation function, followed by a non-dominated sorting genetic algorithm (NSGA-II) with a one-point crossover operator and a heuristic mutation operator. The experiments on MOPSO and NSGA-II are carried out on small, medium, and large scale problems. Moreover, the performance of the two algorithms is compared according to some comparing criteria. The computational results reveal that the NSGA-II performs highly better than the MOPSO algorithm in small scale problems. In the case of medium and large scale problems, the efficiency of the MOPSO algorithm was significantly improved. Nevertheless, the NSGA-II performs robustly in the most important criteria.
Highlights
It is essential for supply chains that customers and manufacturers work together in a coordinated approach in order to achieve information sharing, communication, and high efficiency
The results indicated the superiority of the meta-heuristics algorithms, with the NSGA-III exceeding the multi-objective particle swarm optimization (MOPSO) algorithm in one of the three performance criteria
Since a unique optimal solution cannot be defined in the multi-objective optimization structure, the MOPSO uses a non-dominated solution procedure where particles randomly choose their leaders from an approximated Pareto curve
Summary
It is essential for supply chains that customers and manufacturers work together in a coordinated approach in order to achieve information sharing, communication, and high efficiency. There is a trade-off between the tardiness penalties, the operation time of the entire process (makespan), and the earliness penalties and delivery costs In this regard, simultaneous optimization of this trade-off can make the problem more complicated, but makes it closer to reality. A sequence of processing jobs must be found in order to optimize the total operation time and other supply chain costs using one or more objective functions (Cheng & Sin, 1990; Sivrikaya-Şerifoǧlu & Ulusoy, 1999). Wang and Cheng (2000) performed the first research on parallel machines scheduling with batch delivery aiming at minimizing the total operation time and delivery costs. Minimizing the total weighted number of tardy orders, total earliness penalty, and total batch delivery costs is considered as the second objective function. The last section (Sect.7) draws conclusions and proposes future research directions
Sign-in/Register to view highlights & summary 
Published Version
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call