Research On The Key Problems Of Kinematics And Dynamics Control Based On Underwater Manipulator

At presents,robotics industries,which are widely used in many fields such as medicine,industry,ocean exploration,entertainment,space technology,information and so on,develops very rapidly around the world.The development of underwater robot is especially fast and becomes an independent research system.Due to the particularity of the working environment,understanding manipulators' dynamic character thoroughly,increasing the accuracy of the control and enhancing its work ability have significant sense.This paper analysis the kinematics,dynamics and precise control in complex environment based on a self-designed Four-Degree-of-Freedom underwater manipulator.First of all,coordinate systems of manipulator's link are set up based on the D-H method,then equations of kinematics are derived.Based on that,An innovative method which combines the rotation of singular surface and Fixed-step method is proposed.This method has clearer results comparing with the Monte Carlo method.Because the one-to-many relationship of variable in task and joint space,using Neural Network Intelligent Optimization Algorithms to solve this problem to increase the stability of inverse kinematics.Meanwhile,by using PSO algorithm to optimize the BP neural network model,an algorithm with high precision and fast convergence speed is obtained.Secondly,combined with the Lagrange method and Morison equation,a complete dynamic model of underwater manipulator is established.Next,the dynamic character of this manipulator is analysed,including the inertia and gravitational moment.Through numerical calculation,the influence of hydraulic resistance and buoyancy to manipulator are studied.According to the test and simulation,we propose some methods to analyze the influence of hydrostatic environment on the trajectory and angular velocity of each moving joint,provide dynamic model for design of controller.Thirdly,we design the control algorithm based on the dynamic model of traditional two-freedom rigid link and the manipulator.We presented the computed torque control algorithm and the sliding mode control algorithm based on the computed torque control.To address the issue of chattering problem for sliding mode control and unavailable normalization factor,fuzzy adaptive control theory is added to the previous algorithm.Finally,an adaptive fuzzy sliding mode is proposed.Compared with the traditional algorithm,its stability can be proved simpler and it can adjust the control inputs in a real-time way according to the date,eliminating the chattering of sliding mode control and the interference of complex environment.Finally,the control simulation analysis is carried out on the simulation platform-Simulink based on the dynamic model of manipulator and designed control algorithm.On the one hand,the effects of different control algorithms on trajectory tracking based on the underwater manipulator are analyzed.On the other hand,the SimMechanics simulation platform is used to achieve the visual motion control process of the manipulator,so the feasibility of the proposed control algorithm is verified.