Research On Trajectory Planning Method Of Spatial Hyper-redundant Quadruped Crawling Robot

With the booming development of the space industry,space exploration has become one of the priorities of all countries.The in-orbit service technologies such as assembly,detection and maintenance of giant spacecraft will be an important trend in the future development.Among them,space robots can withstand high-danger and high-load tasks and have broad application prospects in in-orbit services.Space robots need to complete crawling,transportation,assembly,detection and other related operations in the face of in-orbit tasks,and the precision and stable movement of robots is the key to achieve the task.Traditional space robots are mostly single-armed or dual-armed non-redundant robots,and redundant robots are only seven degrees of freedom,which have a small range of movement,low flexibility and poor robustness.In the complex and changeable situation of space tasks,their applicability is very limited.Therefore,it is imperative to develop and research related technologies of space crawling robot.In this paper,the trajectory planning method of nine-degree-of-freedom superredundant quadruped crawling robot was studied,aiming to make the space robot to complete the movement task flexibly,stably,accurately and smoothly on the space truss.The specific research content was included the following aspects:（1）In terms of the modeling problem of spatial hyper-redundant quadruped crawling robot,the screw method was adopted in this paper.The spinor method was equivalent to the combination of rotation and movement of the manipulator around an axis with the help of the related theory of spinor algebra.In addition,Monte Carlo Method was used to analyze the robot’s reachable workspace.In dynamic modeling,the equal efficiency method was used to transform the system relationship between open chain system and closed chain system.This method could describe the whole motion directly,had the characteristics of direct geometric description and simple analysis process,and could effectively lay a foundation for the subsequent research work.（2）Gait design was involved in the movement process of spatial hyper-redundant quadruped crawling robot.In the aspect of gait design,variable coordinate system was designed to analyze the crawling robot’s two states,namely stride leg and body movement,and the quadruped and body movement synchronization strategy was adopted to improve the movement efficiency.At the same time,in order to ensure the robot crawling stability,high efficiency and flexibility,three gait,bipedal gait and single gait were designed to achieve a variety of gait movements.（3）In terms of the path and trajectory planning of the robot foot,aiming at the traversal sequence of the truss task points,a slime mold-ant colony fusion algorithm was designed to achieve the path planning of the robot foot with the shortest distance.Meanwhile,a piecewise polynomial interpolation method based on slime model algorithm was designed to ensure the smooth and optimal time trajectory of the foot in Cartesian space.The slime mold-ant colony fusion algorithm mainly involved: Firstly,according to the "key pipeline key cultivation" method abstracted from the slime mold algorithm about network,the path points were planned and high-quality pipelines are screened out,and both ends of high-quality pipelines were equivalent to fixed point pairs.Then,according to the principle of fixed selection,the quality pipes screened by slime molds were combined with the state transfer formula of ant colony algorithm to complete variable updating and obtain the fusion algorithm.This method effectively reduced the traversal dimension of path points and improves the efficiency of the algorithm.At the same time,the high-quality pipeline could solve the local optimization problem of the algorithm,and effectively improve the robot foot path optimization results.The piecewise polynomial interpolation method based on slime mold algorithm combined the piecewise idea,polynomial interpolation function and slime mold algorithm to complete the design.The method mainly was consisted of selecting three segments of cubic polynomial to fit the trajectory between two points of the robot foot,and optimizing the parameters of the segmented polynomial by using the slime model algorithm.This design method not only ensured the smoothness of the trajectory interpolation,but also reduced the polynomial dimension and complexity,and obtained the time-optimal trajectory interpolation results in cartesian space of the robot foot.（4）In the aspect of trajectory planning of robot joint space,aiming at the problem of inverse solution and trajectory tracking accuracy of redundant robot,PI controller was combined with Henry gas solubility algorithm（HGSO）with strong optimization ability,and PI control method of parameter setting based on gas solubility algorithm was proposed.The algorithm mainly was consisted of three modules: Firstly,the PI controller with simple structure and good effect was selected to deal with the position deviation of the robot end;Secondly,the position deviation was abstracted as the end velocity of the manipulator,and the space velocity of the joint was calculated by Jacobian inverse matrix,and then the joint Angle was obtained by integrating the joint angular velocity.Joint constraints and joint singularity avoidance were realized by weighted Jacobian matrix and pseudo inverse Jacobian matrix;Thirdly,Henley gas solubility algorithm with strong optimization performance was used to set PI controller parameters,and the optimized controller was used to obtain the track tracking results with the minimum error.Finally,the effectiveness of the proposed method was verified by the linear and curvilinear trajectory planning experiments of a nine-DOF manipulator.The experimental results showed that when solving the joint space trajectory tracking problem of 9-DOF redundant manipulator,the joint angles and angular velocities of the optimized trajectory planning results were within the constraint range,and the motion was stable and the tracking accuracy was high.（5）In the crawling process of spatial hyper-redundant quadruped robot,the closed chain movement process of the body moving forward would produce force interaction,so that the body and joints of the robot would produce impact force or torque,affecting the robot movement.Therefore,the variable stiffness control of the robot motion process was designed to make the robot complete the compliant motion.In this paper,a piecewise hybrid impedance control method based on Henry solubility algorithm was designed by combining centralized control with distributed control.The method mainly was included: Firstly,the internal force interaction between the body and the manipulators were analyzed to complete the modeling of equivalent internal force;Then,through the analysis of centralized and distributed impedance controller,the output control torque function formula of each manipulator was obtained,and the control torque was applied to the manipulator to complete the planning of coordinated control.In the process of moving the robot body,a centralized impedance control with simple structure was used in the front part,and a distributed impedance control with high precision was used in the moving part near the target position;Finally,according to the constraints and expected indexes,the parameters of the impedance controller were optimized by the improved Henry gas solubility algorithm,and the appropriate parameters were selected to adjust the dynamic relationship between the robot position and the force to achieve compliant motion.The Henry gas solubility algorithm was improved the optimization ability and speed by designing the bidirectional parallel distribution strategy of the initial gas particles and adding the variation factor strategy when the particles are updated.Thus,it could better serve the design of high efficiency and high precision impedance controller and realize flexible movement of robot with variable stiffness.