Trajectory planning of a 4-RR(SS)2 high-speed parallel robot

The trajectory-planning method for a novel 4-degree-of-freedom high-speed parallel robot is studied herein. The robot’s motion mechanism adopts RR(SS)2 as branch chains and has a single moving platform structure. Compared with a double moving platform structure, the proposed parallel robot has better acceleration and deceleration performance since the mass of its moving platform is lighter. An inverse kinematics model of the mechanism is established, and the corresponding relationship between the motion parameters of the end-moving platform and the active arm with three end-motion laws is obtained, followed by the optimization of the motion laws by considering the motion laws’ duration and stability. A Lamé curve is used to transition the right-angled part of the traditional gate trajectory, and the parameters of the Lamé curve are optimized to achieve the shortest movement time and minimum acceleration peak. A method for solving Lamé curve trajectory interpolation points based on deduplication optimization is proposed, and a grasping frequency experiment is conducted on a robot prototype. Results show that the grasping frequency of the optimized Lamé curve prototype can be increased to 147 times/min, and its work efficiency is 54.7% higher than that obtained using the traditional Adept gate-shaped trajectory.

Real-time quintic Hermite interpolation for robot trajectory execution

This paper presents a real-time joint trajectory interpolation system for the purpose of frequency scaling the low cycle time of a robot controller, allowing a Python application to real-time control the robot at a moderate cycle time. Interpolation is based on quintic Hermite piece-wise splines. The splines are calculated in real-time, in a piecewise manner between the high-level, long cycle time trajectory points, while sampling of these splines at an appropriate, shorter cycle time for the real-time requirement of the lower-level system. The principle is usable in general, and the specific implementation presented is for control of the Panda robot from Franka Emika. Tracking delay analysis is presented based on a cosine trajectory. A simple test application has been implemented, demonstrating real-time feeding of a pre-calculated trajectory for cutting with a knife. Estimated forces on the robot wrist are recorded during cutting and presented in the paper.

Online local modeling and prediction of batch process trajectories using just-in-time learning and LSTM neural network

This paper deals with the online sample trajectory prediction problem of batch processes considering complex data characteristics and batch-to-batch variations. Although some methods have been proposed to implement the trajectory interpolation problem for quality prediction and monitoring applications, the accuracy and reliability are not ensured due to data nonlinearity, dynamics and other complicated feature. To improve the data interpolation performance, an improved JITL-LSTM approach is designed in this work. Firstly, an improved trajectory-based JITL strategy is developed to extract similar local trajectories. Then the LSTM neural network is used on the basis of the extracted trajectories with a modified network structure. Therefore, trajectory prediction and interpolation can be achieved according to the local JITL-LSTM model at each time index. A simulated fed-batch reactor process is presented to demonstrate the effectiveness of the proposed method.

Batting an in-flight object to the target

Striking a flying object such as a ball to some target location is a highly skillful maneuver that a human being has to learn through a great deal of practice. In robotic manipulation, precision batting remains one of the most challenging tasks in which computer vision, modeling, planning, control, and action must be tightly coordinated in a split second. This paper investigates the problem of a two-degree-of-freedom robotic arm intercepting an object in free flight and redirecting it to some target with a single strike, assuming all the movements take place in one vertical plane. Two-dimensional impact is solved under Coulomb friction and energy-based restitution with a proof of termination. Planning combines impact dynamics and projectile flight mechanics with manipulator kinematics and image-based motion estimation. As the object is on the incoming flight, the post-impact task constraint of reaching the target is propagated backward in time, while the arm’s kinematic constraints are propagated forward (via joint trajectory interpolation), all to the pre-impact instant when they will meet constraints that allow batting to happen. All the constraints (16 in total) are then exerted on the arm’s pre-impact joint angles and velocities, which are repeatedly planned based on updated estimates of the object’s motion captured by a high-speed camera. The arm keeps adjusting its motion in sync with planning until batting takes place. Experiments have demonstrated a better batting performance by a Barrett Technology WAM Arm than by a human being without training.