Computational intelligence in robotics and mechatronics

摘要

There are many growing demands to make systems more intelligent to meet more advanced requirements. This paper provides the outline of computational intelligence (CI), the intelligent system architectures and design methods of the system. Intelligent systems can be made by the method of CI, such as fuzzy logic, neural networks and evolutional computation. Furthermore, intelligent systems require the capabilities for evolution, adaptation and learning. This paper discusses the roles of evolution, adaptation and learning on the robotics and mechatronics. Next, we show two models of intelligent systems: hierarchical intelligent system and structured intelligent system.