Motor-Cortex-Like Recurrent Neural Network and Multitask Learning for the Control of Musculoskeletal Systems

The musculoskeletal robot is a promising direction of the next-generation robots. However, current control methods of musculoskeletal robots lack multitask learning ability, great generalization, and biological plausibility. In this article, a motor-cortex-like recurrent neural network (RNN) and a reward-modulated multitask learning method are proposed. First, inspired by the dynamic system hypothesis of motor cortex, the RNN is introduced to transform movement targets into muscle excitations. The condition that makes an RNN generate motor-cortex-like consistent population response is investigated. Second, a reward-modulated multitask learning method of such an RNN is proposed. In the experiments, the control of a musculoskeletal system is realized with multitask learning ability, great generalization, and robustness for noises. Furthermore, the RNN and muscle excitations demonstrate motor-cortex-like consistent population response and human-like muscle synergies, respectively. Therefore, the proposed method has better performance and biological plausibility, and verifies the neural mechanisms in the robotic research.