Generalization of movements in quadruped robot locomotion by learning specialized motion data

Abstract

Machines that are sensitive to environmental fluctuations, such as autonomous and pet robots, are currently in demand, rendering the ability to control huge and complex systems crucial. However, controlling such a system in its entirety using only one control device is difficult; for this purpose, a system must be both diverse and flexible. Herein, we derive and analyze the feature values of robot sensor and actuator data, thereby investigating the role that each feature value plays in robot locomotion. We conduct experiments using a developed quadruped robot from which we acquire multi-point motion information as the movement data; we extract the features of these movement data using an autoencoder. Next, we decompose the movement data into three features and extract various gait patterns. Despite learning only the "walking" movement, the movement patterns of trotting and bounding are also extracted herein, which suggests that movement data obtained via hardware contain various gait patterns. Although the present robot cannot locomote with these movements, this research suggests the possibility of generating unlearned movements.

Generalization-Based Acquisition of Training Data for Motor Primitive Learning by Neural Networks

“…Action representations and Reinforcement Learning (RL) are discussed in, for example, [5,18,30,31]. Furthermore, action generalization in feature space of a quadruped robot is discussed in [32].…”

Learning multiple gaits of quadruped robot using hierarchical reinforcement learning

“…Humans and animals show various gait patterns and smooth transitions depending on the environment and desired velocity. Researchers actively worked on generating and controlling multiple gaits considering the relationship between them [20], [25]. Erden et al designed specific reward based on heuristics and learned adaptive gaits in limited environment [19].…”

Projeto E Acionamento De Um Robô Lagartixa Pneumático Com Aplicação De Aprendizado Computacional

“…Em [32][33][34][35][36] há pesquisas que apresentam discussões sobre o funcionamento de redes neurais para os fins de treinamento de aprendizado de movimento de robôs, além de mostrar aplicações com diversas técnicas.…”

“…Ji et al [35] projetaram um robô quadrúpede que utiliza atuadores não rígidos e treinaram seu gait para funcionar de forma otimizada nas tarefas de movimentação, além de fazerem testes para verificação de funcionamento. Yamamoto et al [36] executaram uma captura de dados de movimentação em um robô quadrúpede, com diversos gaits distintos e os analisaram para achar padrões de movimentação através de aprendizado de máquina.…”