A Deep Reinforcement Learning Based Efficient Optimization Solution Method for Inverse Kinematics of Hyper-redundant Robot | IEEE Conference Publication | IEEE Xplore