In order to meet the needs of spatial obstacle avoidance trajectory planning for the manipulator,a six DOF manipulator obstacle avoidance algorithm based on NMPC was proposed.Compared with the previous obstacle avoidance algorithm,this one allowed trajectory accuracy control while avoiding obstacles.Firstly,an optimal control model of DOF robotic arm autonomous collision prevention was developed.Secondly,the linearization of robotic arm manipulator nonlinear system was used to improve the real-time performance of the system,and the optimal solution was found by cutting and searching.Finally,it could achieve the effect of self-avoidance under the premise of meeting control constrains and trajectory constrains.The algorithm was verified on MATLAB platform.The results show that the method can be used to effectively plan an obstacle-free motion trajectory that meets the performance requirements of the robotic arm,to realize the collision control of the robotic arm in dynamic environment and to improve the operational safety of the robotic arm effectively.