A Multi-Agent System Approach to Train Delay Handling