Fuzzy Linguistic Odor Cognition for Robotics Olfaction

The robotic olfaction helps the robot to measure the odor and provides it the environment recognition ability. This paper proposes a fuzzy linguistic odor cognition method for robotics olfaction. The language is the presentation mode and the information medium of human mind. The core idea of the proposed method is that the fuzzy linguistic rules are used to realize the brain-inspired knowledge representation. The proposed method is a cognitive model which consists of a set of fuzzy linguistic rules. The input variables of the rule antecedents are the signals of different gas sensors on the robotic electronic nose. The rule consequent is the odor label. The fuzzy linguistic rules are initialized by the density-based spatial clustering of applications with noise. The parameters of the membership functions and the rule consequents are estimated by an iterative optimization process. The experiments are performed on two real robotics olfaction data sets obtained under uncontrolled realistic conditions. The proposed method is compared with decision tree, k-nearest neighbor (KNN), support vector machine (SVM), and backpropagation neural network (BPNN). The experimental results verify that the effectiveness of the proposed method is better than that of decision tree, KNN, SVM, and BPNN.