Trajectory Deformation Based On Energy Optimization And Obstacle Avoidance

As a wearable auxiliary device,exoskeleton robots are used in more and more fields,and people’s requirements for safety and efficiency are also increasing.Therefore,it is of great significance to study the trajectory generation method and control strategy in human-computer interaction of exoskeleton robots.Based on exoskeleton robot,this paper proposes a trajectory generation strategy and adaptive control method based on energy optimization equation for human-computer interaction.According to the interaction force applied by the human to the robot,the energy optimization equation is solved through the neural network in real time to generate the reference trajectory of the robot’s movement.The reference trajectory is the trajectory of the robot conforming to the movement of the person,which contains the intention of the person’s movement,and at the same time,the interaction force can be minimized to reduce the operation burden of the person.Taking into account the uncertain information in the robot dynamics model(including unknown mass and moment of inertia,etc.),the corresponding adaptive controller is designed based on the Lyapunov method,so that the robot can track the reference trajectory with high accuracy.The main research contents of this paper are summarized as follows:(1)On the basis of energy optimization and obstacle avoidance,this paper consider the physical limits of the robot joints,and realize the trajectory generation of the robot arms of the two joints through human-computer interaction.Different from the related work,this paper not only considered the energy consumption of the manipulator in the trajectory generation process,but also considered the feasibility of the trajectory,adding an inequality constraint to the original equation constraint to avoid the robot arm exceeding the joint limit.After transforming the optimization problem into a constrained equation,this paper use a neural network to quickly solve the optimal trajectory.(2)An adaptive control framework with two loops is designed.According to the interaction force between man and robot,the outer loop quickly solves the energy optimization equation with constraints through the neural network,and reshapes the robot’s motion reference trajectory,which is used as the input of the inner loop;the inner loop uses the robot’s motion state feedback,An adaptive fuzzy controller with disturbance observer is designed based on Lyapunov stability theory and fuzzy control theory,in which the uncertain information of dynamics is processed based on the regression,and the disturbance observer is used to compensate the system Unknown external interference.This control scheme combines the energy-optimized trajectory generation method,which not only effectively handles the physical constraints of the robot,but also makes the robot’s tracking performance satisfactory.Compared with traditional methods,the designed controller can greatly reduce the interaction force required by humans.