Aggregation operators in accurate potential field building

Abstract

The need of information fusion has become increasingly important in various disciplines of modern engineering and artificial intelligence. Aggregation operators are efficiently support the merge of information and data from different sources in order to make proper decisions or to represent and improve generic knowledge of various system. The range sensing is a foundation of intelligent mobile robotics. Intelligent processing of data obtained by combination of sensors allows extracting useful information to estimate the state of the robot's environment especially by potential field building method. The accuracy of a potential field is based on the distance estimate vector obtained by measurements of the agents. In order to introduce more realistic distance evaluation process we propose the application of the weighted ordered weighted averaging (WOWA) operator in the multi-agent system (MAS). The traditionally used weighting method that required the tuning of a gain factor is replaced with the aggregation operator. The proposed technique allows considering both the importance of measurements and the effects of uncertainties, measurement errors at the scan points. Simulation results validate that the proposed technique improves the accuracy of the built potential field besides applying lower number of agents.