Diffusion Model And Reinforcement Learning