Adaptation, learning and evolution for intelligent robotic system

Living things have close linkage of perception-decision making-action. Based on sensory information from the external environment, living things make decisions and actions for surviving and adapting in dynamic environments. The driving force for adaptation is mainly learning and genetic inheritance (evolution). Intelligent robots also need the adapting ability. The intelligence of a robot emerges as the result of the synthesis of local modules by the simultaneous processing of perception, decision making and action. Consequently, the intelligence depends on the structure for processing information on the hardware and software. Therefore, we propose a robotic system with structured intelligence. A robot with structured intelligence adapts to environment and acquires skill and knowledge through the interaction with the dynamic environment by computational intelligence including neural network, fuzzy system and genetic algorithm