Other

Atr Auditory and Visual Perception Research Laboratories

Advanced Telecommunications Research Institute International

Mitsuo Kawato

Ryoji Suzuki

Jun Morimoto

Noriaki Yahata

Giuseppe Lisi

Kenji Doya

Yoshiharu Maeda

Yoji Uno

The invention provides a novel highly-adaptive agent learning machine comprising a plurality of learning modules each having a set of reinforcement learning system which works on an environment and determines an action output for maximizing a reward provided as a result thereof and an environment predicting system which predicts a change in the environment, wherein a responsibility signal is calculated such that the smaller a prediction error of the environment predicting system of each of the learning modules, the larger the value thereof, and the action output by the reinforcement learning system is weighted in proportion to the responsibility signal, thereby providing an action with regard to the environment. The machine switches and combines actions optimum to various states or operational modes of an environment without using any specific teacher signal and performs behavior learning flexibly without using any prior knowledge.

Agent learning machine

A movement trajectory generating system of a dynamical system uses neural network units (1, 2, 3) including cascade connection of a first layer (11, 21, 31), a second layer (12, 22, 32), a third layer (13, 23, 33) and a fourth layer (14, 24, 34), to learn a vector field of differential equations indicating forward dynamics of a controlled object (4). Conditions concerning trajectories of a final point and a via-point of movement of the controlled object and locations of obstacles are given from a motor center (5). While smoothness of movement is ensured by couplings of electric synapses using errors with respect to those conditions as total energy, least dissipation of energy is attained, whereby trajectory formation and control input for realizing the trajectory are obtained simultaneously.

Movement trajectory generating method of a dynamical system

Growing community of inventors

Kyoto, Japan

Mitsuo Kawato