科研之友
微信
新浪微博
Facebook
分享链接
摘要
This paper proposes a novel grey qualitative based approach to solve the expression problem of uncertain information in cognitive mapping. The proposed approach is based on information theory together with grey correlation analysis to simulate the process of human recognizing the environment. As more data are collected, new line-segments are detected by extracting the feature having large value of entropy, and the existing line-segments are updated by calculating grey correlation degree of the sensing data with them. These line-segments are qualitative characterization of the environment for the agent. The intersection points of the line-segments are optimum in the sense that they are the features yielding the most entropy reduction with existing information. These feature points are named grey qualitative characteristic values which are the basic elements of grey qualitative map. The method has the advantage of calculating the grey qualitative characteristic values to decrease the computation and storage cost in real time implementations. Simulation results show that the grey qualitative based approach for feature extraction is suitable for indoor environment and is effective in simulating human to process environmental information.
