The multi-agent cooperating learning method is proposed to enable the multi-agent system to own quickly the reinforcement learning and BDI ability. The model connects the reinforcement learning with the BDI model.

• 将Agent的独立强化学习与BDI模型结合起来；使多Agent系统不但拥有强化学习的高度反应性和自适应性；而且拥有BDI的推理能力；使只使用数值分析而忽略推理环节的强化学习结合了逻辑推理方法.