Step climbing method for crawler type rescue robot using reinforcement learning with Proximal Policy Optimization | IEEE Conference Publication | IEEE Xplore