A Probabilistic Binary Multi-Unit Combinatorial Reverse Auction Problem and an Imperialist Competitive Algorithm for its Solution

Abstract

Winner determination problem in combinatorial auctions is a useful concept in different areas such as e-business. In this paper, a probabilistic binary multi-unit combinatorial reverse auction problem as a version of reverse auctions has been modeled, in which the auctioneer intends to make an initial decision about the demand allocation by considering the previous proposals into account. In this proposed model, the number of items supplied by suppliers and its proposed price are stochastic variables. An imperialist competitive algorithm as a population based metaheuristic algorithm has been used to solve this problem. The results have shown that the imperialist competitive acts more efficiently and effectively in comparison with the genetic algorithm, also it attains the near optimal solution in a reasonable time. This paper has prepared an efficient approach for large company dealt with several proposals sent by different suppliers.