Constrained Sampling-based Trajectory Optimization using Stochastic Approximation

TL;DR

This paper introduces a sampling-based trajectory optimization method for constrained systems, effectively handling control and state constraints with improved solution quality and faster convergence demonstrated on cartpole and quadcopter simulations.

Contribution

It extends stochastic search techniques to incorporate box control constraints and nonlinear state constraints using truncated distributions and penalty functions.

Findings

Outperforms previous methods in solution quality

Achieves faster convergence in constrained optimization

Successfully applied to cartpole and quadcopter systems

Abstract

We propose a sampling-based trajectory optimization methodology for constrained problems. We extend recent works on stochastic search to deal with box control constraints,as well as nonlinear state constraints for discrete dynamical systems. Regarding the former, our strategy is to optimize over truncated parameterized distributions on control inputs. Furthermore, we show how non-smooth penalty functions can be incorporated into our framework to handle state constraints. Simulations on cartpole and quadcopter show that our approach outperforms previous methods on constrained sampling-based optimization, in terms of quality of solutions and convergence speed.