Method for controlling a robot and robot controller

Abstract

According to various embodiments, a method for controlling a robot using control parameter values from a non-Euclidean original control parameter space is described, the method comprising performing a Bayesian optimization of an objective function representing a desired control objective of the robot over the original control parameter space; and controlling the robot in accordance with a control parameter value from the original control parameter space found in the Bayesian optimization, wherein the Bayesian optimization includes: transforming the original control parameter space to a reduced control parameter space using the observed control parameter values, wherein the original control parameter space comprises a first number of dimensions, the reduced control parameter space comprises a second number of dimensions, and the first number of dimensions is higher than the second number of dimensions; determining an evaluation point of the objective function in the reduced control parameter space by searching an optimum of an acquisition function in an iterative search, comprising, in each iteration, updating a candidate evaluation point using a search direction in the tangent space of the reduced control parameter space at the candidate evaluation point; mapping the updated candidate evaluation point from the tangent space to the reduced control parameter space; and using the mapped updated candidate evaluation point as evaluation point for a next iteration until a stop criterion is fulfilled; and mapping the determined evaluation point from the reduced control parameter space to the original control parameter space.

Description

technical field [0001] The present disclosure relates to a method and a robot controller for controlling a robot. Background technique [0002] Bayesian optimization (BO) has recently become popular in robotics due to its data-efficient and gradient-free approach to optimize control parameters and parametric policies in direct reinforcement learning. [0003] However, the performance of Bayesian optimization algorithms for optimizing robot control parameters may be severely affected when the parameter space is high-dimensional. Furthermore, it should be noted that the parameter space in robot control can be a non-Euclidean space. For example, stiffness can be represented by a positive definite matrix (i.e., the parameter space forms a manifold of symmetric positive definite matrices), and vectors on the unit sphere can be used to represent orientations. [0004] In view of the above, efficient methods for performing Bayesian optimization on high-dimensional non-Euclidean p...

AI Technical Summary

Problems solved by technology

[0003] However, the performance of Bayesian optimization algorithms for optimizing robot control parameters may be severely affected when the parameter space is high-dimensional

Images