This webpage supplements to the research paper:
Qian Hu*, Lijun Wei, Andrew Lim. The two-dimensional vector packing problem with general costs. Omega-International Journal of Management Science. 2018. 74: 59-69. DOI
The two-dimensional vector packing problem with general costs (2DVPP-GC) arises in logistics where shipping items of different weight and volume are packed into cartons before being transported by a courier company. In practice, the delivery cost of a carton of items is usually retrieved from a cost table. The costs may not preserve any known mathematical function since it could specify arbitrary price at any possible weight. Such a general pricing scheme meets a majority of real-world bin packing applications, where the price of delivery service is determined by many complicated and correlated factors. Compared to the classical bin packing problem and its variants, the 2DVPP-GC is more complex and challenging. To solve the 2DVPP-GC with minimizing the total cost, we propose a memetic algorithm to compute solutions of high quality. Fitness functions and improved operators are proposed to achieve effectiveness. Computational experiments on a variety of test instances show that the algorithm is competent to solve the 2DVPP-GC. In particular, optimal solutions are found in a second for all the test instances that have a known optimal solution.
Our algorithms were implemented as a sequential algorithm in Java. The experimental results were obtained on a workstation equipped with an Intel(R) Xeon(R) E5-1607 CPU clocked at 3.10 gigahertz, 8 gigabytes RAM and running Windows 10. The test data and results are available online at \url{https://sites.google.com/site/orlib222/2dvpp-gc}.
2DVPP-GC test instances: https://drive.google.com/open?id=0B7azD5PrlaaNMExrRGxOTFZ3cGM
Results: https://drive.google.com/open?id=0B7azD5PrlaaNRzFpX1hOZGpEdTg