Deep reinforcement learning for tensegrity robot locomotion | IEEE Conference Publication | IEEE Xplore