Memory of Motion for Warm-Starting Trajectory Optimization

Lembono, Teguh Santoso; Paolillo, Antonio; Pignat, Emmanuel; Calinon, Sylvain

doi:10.1109/lra.2020.2972893

Cited by 39 publications

(27 citation statements)

References 22 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The problem of inferring the warm start is formulated as a regression problem g(θ) = y where the input task consists of the current robot state and goal location, and the output y consists of the corresponding state and control trajectories. Such a memory of motion [27], [38] has been shown to improve the convergence of DDP significantly. It relies on a dataset of optimal trajectories built off-line and encoded by machine learning.…”

Section: A Building the Memory Datasetmentioning

confidence: 99%

See 1 more Smart Citation

Whole Body Model Predictive Control with a Memory of Motion: Experiments on a Torque-Controlled Talos

Dantec

Budhiraja

Roig³

et al. 2021

2021 IEEE International Conference on Robotics and Automation (ICRA)

Self Cite

View full text Add to dashboard Cite

This paper presents the first successful experiment implementing whole-body model predictive control with state feedback on a torque-control humanoid robot. We demonstrate that our control scheme is able to do whole-body target tracking, control the balance in front of strong external perturbations and avoid collision with an external object. The key elements for this success are threefold. First, optimal control over a receding horizon is implemented with Crocoddyl, an optimal control library based on differential dynamics programming, providing state-feedback control in less than 10 ms. Second, a warm start strategy based on memory of motion has been implemented to overcome the sensitivity of the optimal control solver to initial conditions. Finally, the optimal trajectories are executed by a low-level torque controller, feedbacking on direct torque measurement at high frequency. This paper provides the details of the method, along with analytical benchmarks with the real humanoid robot Talos.A video of the experiment is available at https://peertube.laas.fr/videos/watch/cbc25927-337c-4635-a1bc-153b9aeb4135

show abstract

Section: A Building the Memory Datasetmentioning

confidence: 99%

“…Yet the task considered in this work is multimodal, i.e., there are several qualitatively different trajectories to achieve one task. For such problem, GPR performs poorly as observed in [38], hence we choose NN in this work instead.…”

Section: A Building the Memory Datasetmentioning

confidence: 99%

Whole Body Model Predictive Control with a Memory of Motion: Experiments on a Torque-Controlled Talos

Dantec

Budhiraja

Roig³

et al. 2021

2021 IEEE International Conference on Robotics and Automation (ICRA)

Self Cite

View full text Add to dashboard Cite

show abstract

“…1) Learning Single-step Motion: Following our previous work [23], we formulate the problem of learning the singlestep motion as a regression problem to approximate the mapping f : x → y, where x is the task and y is the corresponding trajectory output. We separate the database into left-leg and right-leg movements, as this yields better results than combining both and let the memory learns how to choose the leg.…”

Section: Learning Strategiesmentioning

confidence: 99%

Learning How to Walk: Warm-starting Optimal Control Solver with Memory of Motion

Lembono¹,

Mastalli²,

Fernbach³

et al. 2020

Preprint

Self Cite

View full text Add to dashboard Cite

In this paper, we propose a framework to build a memory of motion for warm-starting an optimal control solver for the locomotion task of a humanoid robot. We use HPP Loco3D, a versatile locomotion planner, to generate offline a set of dynamically consistent whole-body trajectory to be stored as the memory of motion. The learning problem is formulated as a regression problem to predict a single-step motion given the desired contact locations, which is used as a building block for producing multi-step motions. The predicted motion is then used as a warm-start for the fast optimal control solver Crocoddyl. We have shown that the approach manages to reduce the required number of iterations to reach the convergence from ∼9.5 to only ∼3.0 iterations for the singlestep motion and from ∼6.2 to ∼4.5 iterations for the multi-step motion, while maintaining the solution's quality.

show abstract

“…The research has been supported by the Deutsche Forschungsgemeinschaft (DFG, German Research Foundation) under Germany's Excellence Strategy. 1 Learning & Intelligent Systems Lab, TU Berlin, Germany 2 BITS Pilani, India the initial trajectory [21], [19], [10]. Hence, these methods usually work well in environments with few obstacles or when provided with good initial guesses [20], e.g.…”

Section: Introductionmentioning

confidence: 99%

BITKOMO: Combining Sampling and Optimization for Fast Convergence in Optimal Motion Planning

Kamat¹,

Ortiz-Haro²,

Toussaint³

et al. 2022

Preprint

View full text Add to dashboard Cite

Optimal sampling based motion planning and trajectory optimization are two competing frameworks to generate optimal motion plans. Both frameworks have complementary properties: Sampling based planners are typically slow to converge, but provide optimality guarantees. Trajectory optimizers, however, are typically fast to converge, but do not provide global optimality guarantees in nonconvex problems, e.g. scenarios with obstacles. To achieve the best of both worlds, we introduce a new planner, BITKOMO, which integrates the asymptotically optimal Batch Informed Trees (BIT*) planner with the K-Order Markov Optimization (KOMO) trajectory optimization framework. Our planner is anytime and maintains the same asymptotic optimality guarantees provided by BIT*, while also exploiting the fast convergence of the KOMO trajectory optimizer. We experimentally evaluate our planner on manipulation scenarios that involve high dimensional configuration spaces, with up to two 7-DoF manipulators, obstacles and narrow passages. BITKOMO performs better than KOMO by succeeding even when KOMO fails, and it outperforms BIT* in terms of convergence to the optimal solution.

show abstract

Memory of Motion for Warm-Starting Trajectory Optimization

Cited by 39 publications

References 22 publications

Whole Body Model Predictive Control with a Memory of Motion: Experiments on a Torque-Controlled Talos

Whole Body Model Predictive Control with a Memory of Motion: Experiments on a Torque-Controlled Talos

Learning How to Walk: Warm-starting Optimal Control Solver with Memory of Motion

BITKOMO: Combining Sampling and Optimization for Fast Convergence in Optimal Motion Planning

Contact Info

Product

Resources

About