My goal is to enhance the robustness of multi-agent reinforcement learning systems, enabling them to withstand disruptive factors from dynamic environments. To achieve this, I am developing algorithms and adaptive strategies that allow agents to efficiently adapt to unexpected changes and external influences. My research places particular emphasis on ensuring system reliability under various environmental challenges, such as technical failures or unforeseen human factors. Ultimately, my aim is to create multi-agent reinforcement learning models that provide stable and scalable solutions across complex application domains.