STOMP: Stochastic Trajectory Optimization for Motion Planning¶
Authors: Mrinal Kalakrishnan, Sachin Chitta, Evangelos Theodorou, Peter Pastor, Stefan Schaal
Published: 2011 (Conference Paper)
Source: IEEE International Conference on Robotics and Automation (ICRA)
Algorithm: STOMP
DOI: 10.1109/ICRA.2011.5980280
Summary¶
STOMP improves a trajectory by rolling out a large batch of noisy samples and computing a weighted average (based on a Boltzmann distribution i.e. exponentiated negative with a temperature parameter) over their costs. Effectively a path-space analogue of MPPI. Does not require cost gradients. Noise helps STOMP jiggle out of local minima that CHOMP can get stuck in, a similar mechanism and phenomenon as in general stochastic/perturbed gradient descent (c.f. "How to Escape Saddle Points Efficiently" by Jin et al, 2017 https://proceedings.mlr.press/v70/jin17a.html).
Abstract¶
We present a new approach to motion planning using a stochastic trajectory optimization framework. The approach relies on generating noisy trajectories to explore the space around an initial (possibly infeasible) trajectory, which are then combined to produced an updated trajectory with lower cost. A cost function based on a combination of obstacle and smoothness cost is optimized in each iteration. No gradient information is required for the particular optimization algorithm that we use and so general costs for which derivatives may not be available (e.g. costs corresponding to constraints and motor torques) can be included in the cost function. We demonstrate the approach both in simulation and on a mobile manipulation system for unconstrained and constrained tasks. We experimentally show that the stochastic nature of STOMP allows it to overcome local minima that gradient-based methods like CHOMP can get stuck in.
Links¶
Primary
Standard
Alternate
Tags¶
-
Motion planning
-
Path planning
-
Path optimization
-
Stochastic optimization
-
Path integral
-
Sampling
-
STOMP