A Self-Adaptive Intelligent Control System with Hierarchical Architecture.

This paper presents an Intelligent Control System with Hierarchical Architecture, namely ICSHA. The advantages of three-layered ICSHA are able to carry out multiple tasks and to adjust its control rules automatically to adapt environments. The first layer, Planning Layer, we propose the ASGO (an extension version of Ant System with Genetic Operators) to determine the executing order of subtasks. The visiting schedule is then supervised by second layer, Executive Layer. The third layer, Behavior Layer, is to execute each subtask by using the proposed intelligent control module. While the control module cannot adapt the environment, the proposed iQGA (an improved Quantum Genetic Algorithm) is activated to explore better control actions for producing adaptable control rules. An application of robotic trash collection task is constructed to demonstrate the proposed methods. The simulation results showed that the performances of ASGO and iQGA are satisfied. Simulation result also reveals the adaptability of ICSHA.