Hierarchical hybrid neuromorphic control for robotic motions: Sensing, recognition, planning, adaptation, and learning

Abstract

The authors present a new scheme for intelligent control of robotic manipulators. This control system is analogous to the human cerebral control system. It is a hybrid system of neuromorphic control and symbolic control that includes a neural network for servo control and knowledge-based approximation. The neural network at the servo control level is used for numerical manipulation, while the knowledge-based component is used for the symbolic manipulation. In neuromorphic control, the neural network compensates for the nonlinearity of the system and the uncertainty in the environment. The knowledge base component makes the control strategy in a symbolical manner for the servo level. Simulation and experimental results are included.