An activity-based architecture for intelligent vehicle navigation

Intelligent vehicles must have a proper architecture in order to work in complicated and dynamic environment. After considering the advantages and disadvantages of the two typical architectures, this paper presents an activity-based control architecture for intelligent vehicle navigation. This architecture uses a blackboard to coordinate the interaction between modules, and classifies the activities of the intelligent vehicle into reflecting activity and voluntary activity. The reflecting activity is activated by data and the voluntary activity is selected by the blackboard. The navigation task is accomplished by sequencing these activated activities. The theoretic analyses and simulation results show that this architecture has good intelligence and reactivity.