• Corpus ID: 233423287

Fast Joint Space Model-Predictive Control for Reactive Manipulation

@article{Bhardwaj2021FastJS,
  title={Fast Joint Space Model-Predictive Control for Reactive Manipulation},
  author={Mohak Bhardwaj and Balakumar Sundaralingam and Arsalan Mousavian and Nathan D. Ratliff and Dieter Fox and Fabio Ramos and Byron Boots},
  journal={ArXiv},
  year={2021},
  volume={abs/2104.13542}
}
Sampling-based model predictive control (MPC) is a promising tool for feedback control of robots with complex and non-smooth dynamics and cost functions. The computationally demanding nature of sampling-based MPC algorithms is a key bottleneck in their application to high-dimensional robotic manipulation problems. Previous methods have addressed this issue by running MPC in the task space while relying on a low-level operational space controller for joint control. However, by not using the… 

MPC with Sensor-Based Online Cost Adaptation

Experiments with a wide variety of reaching tasks on an industrial robot manipulator demonstrate that the model predictive control scheme introduced can efficiently solve complex non-convex problems with high- dimensional visual sensory inputs, while still being robust to external disturbances.

Adaptive Model Predictive Control by Learning Classifiers

This paper proposes an adaptive MPC variant that automatically estimates control and model parameters by leveraging ideas from Bayesian optimisation and the classical expected improvement acquisition function and demonstrates the approach on classical control problems under model uncertainty and robotics manipulation tasks.

Autonomous Navigation of AGVs in Unknown Cluttered Environments: log-MPPI Control Strategy

The applicability of log-MPPI for performing a 2D grid-based collision-free navigation in an unknown cluttered environment is demonstrated, showing its superiority to be utilized with the local costmap without adding additional complexity to the optimization problem.

Implicit Distance Functions: Learning and Applications in Control

A novel approach to learn an implicit, differentiable distance function for arbitrary configurations of a robotic manipulator used for reactive control is described, using GPU processing to efficiently query the learned collision representation and obtain an implicit distance between the robot and the environment.

Learning to Optimize in Model Predictive Control

This work focuses on learning to optimize more effectively within Model Predictive Control to improve the update rule within MPC, and demonstrates that this approach can outperform a MPC controller with the same number of samples.

Sequence-of-Constraints MPC: Reactive Timing-Optimal Control of Sequential Manipulation

A model predictive control approach dedicated to robustly execute a single sequence of constraints, which corresponds to a discrete decision sequence of a TAMP plan, which can account for long-term interdependencies of constraints and reactively plan for a timing-optimal transition through all constraints.

Neural Motion Fields: Encoding Grasp Trajectories as Implicit Value Functions

Neural Motion Fields is presented, a novel object representation which encodes both object point clouds and the relative task trajectories as an implicit value function parameterized by a neural network and allows to perform grasping reactively by leveraging sampling-based MPC to optimize this value function.

Model Predictive Control for Dynamic Cloth Manipulation: Parameter Learning and Experimental Validation

—Robotic cloth manipulation is a relevant chal- lenging problem for autonomous robotic systems. Highly deformable objects as textile items can adopt multiple configu- rations and shapes during their

Multi-Objective Reactive Motion Planning in Mobile Manipulators

This thesis employs a novel motion generation framework, Composable Energy Policies (CEP) in mobile manipulators given a prior map generated by Rapid Random Search Tree (RRT), and finds the method can perform 2D planning and whole body control well.

Model Predictive Control for Fluid Human-to-Robot Handovers

This paper proposes to generate smooth motions via an efficient model-predictive control (MPC) framework that integrates perception and complex domain-specific constraints into the optimization problem, and introduces a learning-based grasp reachability model to select candidate grasps which maximize the robot's manipulability, giving it more freedom to satisfy these constraints.

References

SHOWING 1-10 OF 52 REFERENCES

Real-Time Nonlinear Model Predictive Control of Robots Using a Graphics Processing Unit

A new solution method is presented which addresses computational concerns related to nonlinear MPC called nonlinear Evolutionary MPC (NEMPC), and then it is compared to several existing methods to demonstrate that this method is tractable for real-time control of high degree of freedom systems.

An integrated system for real-time model predictive control of humanoid robots

An integrated system based on real-time model-predictive control (MPC) applied to the full dynamics of the robot, which is possible due to the speed of the new physics engine (MuJoCo), the efficiency of the trajectory optimization algorithm, and the contact smoothing methods developed for the purpose of control optimization.

Parameterized and GPU-Parallelized Real-Time Model Predictive Control for High Degree of Freedom Robots

This work presents and evaluates a novel input parameterization method which improves the tractability of model predictive control (MPC) for high degree of freedom (DoF) robots and shows that the parameterized convex solver MPC has faster solve times than traditional MPC for high DoF cases while still achieving similar performance.

Operational Space Control: A Theoretical and Empirical Comparison

An extensive empirical results demonstrate that one of the simplified acceleration-based approaches can be advantageous in terms of task performance, ease of parameter tuning, and general robustness and compliance in the face of inevitable modeling errors.

Whole-body model-predictive control applied to the HRP-2 humanoid

This paper implemented a complete model-predictive controller and applied it in real-time on the physical HRP-2 robot, the first time that such a whole-body model predictive controller is applied in real time on a complex dynamic robot.

Data Efficient Reinforcement Learning for Legged Robots

A model-based framework for robot locomotion that achieves walking based on only 4.5 minutes of data collected on a quadruped robot is presented, which is more than an order of magnitude more sample efficient than current model-free methods.

Full-Body Optimal Control Toward Versatile and Agile Behaviors in a Humanoid Robot

An optimal control framework that takes the full-body dynamics of a humanoid robot into account is developed and a biologically inspired controller maintains the robot's posture for a very short control period.

Real-time model predictive control for quadrotors

The proposed approach combines the standard hierarchical control paradigm that separates the control into low-level motor control, mid-level attitude dynamics control, and a high-level trajectory tracking with a model predictive control strategy to obtain a linear system that models force controlled position and trajectory tracking for the quadrotor.

RelaxedIK: Real-time Synthesis of Accurate and Feasible Robot Arm Motion

We present a real-time motion-synthesis method for robot manipulators, called RelaxedIK, that is able to not only accurately match end-effector pose goals as done by traditional IK solvers, but also

Robust trajectory optimization under frictional contact with iterative learning

A method to optimize trajectories with large modeling errors using a combination of robust optimization and parameter learning and an iterative learning method that uses feedback from execution to learn confidence bounds on modeling parameters is presented.
...