Trajectory Planning of Deep-Sea Hydraulic Manipulator in Joint Space with Flow Constraints

For the grasping task of the deep-sea hydraulic manipulator, it is necessary to carry out kinematic analysis and point-to-point joint motion planning. To reduce the influence of the structural error of the manipulator caused by manufacturing, assembly, transportation, environment, and other factors on the kinematics parameters of the manipulator, RBF neural network is used to identify the kinematics model parameters of the manipulator to obtain more accurate modified D-H parameters. Based on this model and flow constraints, the PSO algorithm is used to plan the joint space trajectory of the deep-sea hydraulic manipulator. The experimental results show that the method can well meet the needs of joint-space planning of the deep-sea hydraulic manipulator that is subjected to flow constraints and lay a foundation for subsequent complex tasks.