Research On Inverse Kinematics Optimization Algorithm Of Underwater Manipulator Based On Differential Evolution

The ocean is rich in mineral and biological resources and is one of the important places for resource exploitation and scientific investigation.Underwater manipulator is necessary equipment for underwater operations,which is widely used in various submarine operations such as resource exploitation,marine scientific research,and military applications.With the continuous expansion and deepening of China’s marine resources development and Marine science research,the independent research and development of deep-sea manipulator becomes more and more important.In the design of manipulator,kinematics analysis is a basic and important part and provides support for subsequent trajectory planning,dynamics analysis,and manipulator control.Due to the constraints of the manipulator configuration(Pieper criterion must be met)and the influence of manufacturing errors,mechanical vibration and friction(the kinematic model of the manipulator deviates from the theoretical model),the traditional method is not suitable for the inverse solution of the manipulator with any configuration.Differential evolution algorithm is one of the most representative and best performing algorithms in evolutionary algorithms.It has the advantages of simple operation,high robustness and strong global convergence.It is suitable for solving optimization problems in engineering field.More and more scholars propose to apply various heuristic algorithms to solve the inverse kinematics of manipulator,which provides a new idea for finding a universal inverse kinematics solution method.After comparing several evolutionary algorithms,this paper selects differential evolutionary algorithm as the basic algorithm to solve the inverse kinematics of manipulator.This paper takes the self-designed underwater seven-function manipulator as the research object and uses traditional methods and evolutionary algorithms to solve the problem of inverse kinematics of the manipulator.Firstly,the kinematics of the self-designed underwater seven-function manipulator is analyzed in this paper.Then the traditional differential evolution algorithm is improved and used to solve the inverse kinematics of the manipulator.Finally,INSDE with elite strategy is proposed and used to solve the inverse kinematics problem of the manipulator.Finally,a non-dominated sorting multi-objective differential evolutionary algorithm INSDE with elite strategy is proposed and used for the solution of the inverse kinematics problem of the robotic arm.It is realized that multiple sets of inverse solutions of the robotic arm are obtained simultaneously in one operation,providing more options for robotic arm trajectory planning,motion control and other aspects.The main research work and innovation points of this paper are as follows:1.The mechanical structure of the manipulator is designed according to the subject task.Furthermore,we perform kinematics solution,analyze velocity characteristic and make trajectory planning.Firstly,the mechanical structure of the underwater manipulator is designed and the main parameters of the manipulator system are determined.Then the inverse kinematic analysis of the robot arm is carried out by algebraic and numerical methods.Based on this,the velocity analysis of the robotic arm is performed to provide a reference basis for the subsequent hydraulic drive component selection.Finally,the accuracy of the inverse solution of the manipulator is verified by trajectory planning.The experimental results show that the changes of joint angle,angular velocity and cross acceleration are smooth and continuous,and the manipulator moves smoothly.2.We propose an inverse solution method of series manipulator based on improved adaptive variational differential evolution algorithm.Aiming at the imbalance of pose error in the inverse solution of the manipulator by intelligent algorithm,this paper introduces the position error under the current iteration times into the pose weight coefficient and realizes that the pose error is always in the same order of magnitude through segmented processing,which improves the stability of the algorithm and can better adapt to the manipulator with different rod lengths.Based on the classical DE algorithm,the adaptive mutation operator(AMO)and the error-related adaptive scaling factor(AF)are proposed to better balance the exploration and development ability of the differential evolution algorithm,which can make the algorithm quickly approach the optimal solution when solving the inverse kinematics problem of the manipulator,avoid the premature convergence of the algorithm,and have stronger global optimization ability.Experiments show that the proposed inverse solution method is suitable for 6-DOF and 7-DOF series manipulators at the same time,and has the advantages of high inverse solution accuracy and good stability.In addition,by reducing the search space and inverse solving the random points in the feasible region,the stability and accuracy of the algorithm are further improved,which provides a new research idea for the general inverse solution method of the series manipulator.3.A non-dominated sorting multi-objective differential evolution algorithm INSDE with elite strategy is proposed.Firstly,the INSDE algorithm adopts adaptive control parameters to better balance the exploratory and development capabilities of the algorithm.Then,the typical congestion distance calculation method is updated to the harmonic average distance calculation method to improve the population distribution and the uniformity of solution set distribution.Finally,in the elite selection stage,a more reasonable cyclic crowding sorting strategy is adopted to prune the population to improve the diversity of solutions.Using performance indexes Inverted Generational Distance(IGD)and Spatial diversity index(Spacing),four algorithms NSGA II,SPEA2,MOEAD,and INSDE are compared on twoobjective and three-objective test functions.Simulation analysis verifies the effectiveness of INSDE algorithm.Finally,the INSDE algorithm with parallel ripple communication strategy is used in the solution of the manipulator’s inverse kinematics problem.The simulation results show that the algorithm can obtain multiple sets of inverse solutions of the manipulator at one time.