Learning, adaptation and evolution for intelligent system

There are many growing demands for making systems intelligent, by which people can cope with the system complexities and software development. The intelligent system must have the capabilities, in principle, for learning, adaptation and evolution, so that the system can adapt to the change of environments, tasks, and systems themselves. This paper provides the foundation and methodologies for the learning, adaptation and evolution, by neural network, fuzzy system and genetic algorithm. Those methods can be applied for various optimization of design, and scheduling problems in automation systems.