Combinatorial double auctions provide an efficient mechanism for resource allocation, maximizing social welfare without requiring private information from participants. However, depending on specific auction conditions, the number of rounds required for convergence can be significantly high. In this work, we extend a previously developed DEVS-based auction model to study the dynamics of social welfare evolution throughout the auction process. While the original model identifies the allocation that maximizes social welfare, our extension enables tracking the proportion of the optimal social welfare achieved at each round. This analysis provides insights into the auction's convergence behavior and serves as the foundation for developing an approach that helps auctioneers select appropriate parameters to achieve a sufficiently good outcome within a reasonable number of rounds in real-world scenarios.
