关键词：传感器；信息造粒理论；数据分析；颗粒状逻辑
摘 要：The main components of Zadeh's scientific contribution are shown. In particular, his information granulation theory is considered in the context of sensor data mining. The concept of granule and granular structures are presented, mereological approach to generating granules is developed, some ways of constructing and using granules in multi-valued logics are shown. Basic ideas of granular computing are discussed. Here the problem of sensor data mining is considered from the viewpoint of granular computing. The notion of logical pragmatics is clarified. The investigation of mathematical structures for data pragmatic analysis is associated with interpreting multi-valued logics and bilattices as formal pragmatics techniques. The concept of Belnap's sensor is proposed, the combinations of Belnap's sensors are discussed. Logical-algebraic models of communication between two sensors are constructed on the basis of bilattices. As a result, the map for dealing with different sensor data inconsistencies is introduced and algorithm for eliminating faults in sensor networks is suggested.