Research On Control Performance Of Underwater Seven-Axis Series Redundant Manipulator

China is abundant in marine resources.In the process of exploitation,utilization and protection of resources,underwater robot play a huge role,and the manipulator it carries is the key to ensure successful operation.With the development of science and technology,marine exploitation is also developing in depth into industries such as marine artificial breeding,comprehensive utilization of marine energy,exploration of seabed resources and expansion of marine space.This makes the tasks performed to be more and more complicated,and the requirements for the flexibility and accuracy of the manipulator are also increasing.In the face of tasks such as exploration and sampling of resources under complex terrain,the construction of underwater platforms,and precise docking of pipelines,the traditional manipulator has low operating efficiency,and even is unable to fulfill the operating requirements.Therefore,it is urgent to study redundant manipulator.The underwater series redundant manipulator is a complex system with redundant degrees of freedom,strong nonlinearity and strong coupling.The research on its control characteristics includes kinematic modeling and inverse kinematic solution,dynamic model and optimization considering the water environment,trajectory planning of compound targets and tracking control of dynamic systems.While the degree of freedom redundancy increases flexibility,it also increases the difficulty of manipulator kinematics,dynamics modeling and solving.Although the tandem structure arrangement is easy to expand the manipulator,the traditional algebraic geometric method cannot be used because it cannot meet the specific configuration requirements.A general inverse kinematics solving method need to be designed.Underwater manipulators operate in complex and variable water environments and are affected by multiple dynamics,making accurate dynamic modeling very difficult.Unlike on land,underwater robots have limited energy carrying capacity.Therefore,the optimal trajectory planning of time and energy composite targets is required.In addition,the control system is the brain of the manipulator,which directly determines the operating accuracy and efficiency of the manipulator.It is a major challenge in the implementation of control strategies how to realize the efficient control of the nonlinear,strong coupling,and dynamic multivariable manipulator system without an accurate mathematical model.This article takes "YL-101" underwater seven-axis series redundant manipulator as the research object,and is dedicated to solving the common key issues in kinematics,dynamics,trajectory planning and traj ectory tracking control.This article discusses and proposes a feasible plan to make some contributions to the innovative theoretical research and extensive use of underwater redundant manipulators.In terms of kinematics,based on the kinematics model of the manipulator,a new improved singular robust inverse method is proposed for solving the problem of inverse kinematics,the traditional singular robust inverse method iteratively exists that the Jacobian matrix approximation and damping factor cannot be optimized in real time,and the stability condition based on Lyapunov function is given.Combined with the particle swarm optimization algorithm,the relational data set between the damping factors and the Jacobian singular values is obtained.After training by neural network,the real-time optimal prediction model of the damping factor is used to obtain the special solutions of inverse kinematics.Combined with the gradient projection method,the general solution of inverse kinematics is obtained and compared with the traditional method.The results show that the proposed new method,while realizing the automatic prediction of the damping factor,has greatly improved in terms of terminal error,joint angle norm,and minimum singular value of the Jacobian matrix.In terms of dynamics,the generalized velocity is introduced to derive and establish the Kane method manipulator dynamic model;according to the actual effective area in contact with water,the cuboid link model of the manipulator is simplified;on this basis,combined with the hydrodynamic formula and existing empirical data,this paper has calculated the main hydrodynamic terms such as water resistance,additional mass force,buoyancy,etc.,and added them to the Kane dynamic equation to construct a complete dynamic model that considers the influence of hydrodynamics,and provides the initial model for trajectory tracking control,and ADAMS simulation verifies the relative effectiveness of the model.In terms of trajectory planning,with the diversity of the space tasks of the manipulator,the optimal trajectory planning researches of joint space,operation space and compound target are carried out,which laid a theoretical foundation for the execution of different tasks.In addition,aiming at the problem of the limited energy carried by the underwater manipulator,according to the kinematics and dynamic equations,non-uniform rational B-splines interpolation is used to perform the optimal trajectory planning of time-energy mixing,and through the natural selection of particle swarm method,The optimal trajectory is solved.The results show that the obtained planning trajectory is smooth,continuous and non-abrupt,which satisfies the requirements of the path and constraints.In terms of tracking control,a genetic fuzzy PD controller is designed to realize the adaptive tracking control of the underwater seven-axis manipulator with no no accurate dynamic model can be obtained in the water environment,and the control law of the manipulator system based on the calculated moment method was given.Based on the ADAMS dynamic model,combined with kinematics and dynamic solution,the ADAMS-Simulink co-simulation model based on the proposed controller is constructed,and the time-energy optimal trajectory was simulated and tracked.The results show that compared with the traditional PD control and torque direct drive control,the proposed genetic fuzzy PD controller has better adaptability,can achieve higher tracking accuracy,and can basically eliminate the real-time error generated in the tracking process,and without error accumulation,the tracking effect performs better.