Research On The Intelligent Information Management Mechanism For The Agricultural Internet Of Things Based On Ontology

The agricultural information intelligent management is an important branch and hotspot of the research on agricultural informationization. The Agricultural Internet of Things (AlOT) is the development and application of the Internet of Things (IOT) in agriculture. Under the background of AIOT, the humans, machines and things are completely connected with one another for intelligent interaction. So the development and application of IOT will bring opportunities that have never been seen before to agricultural information management. However, great challenges will also be brought to us. On the one hand, AIOT information is often characterized with its conspicuous uncertainty due to physical property limitations of AIOT and the dynamic changing of the outer environment, on the other hand, AIOT information is heterogeneous in representation because of the open-end, distributivity and heterogeneity of AIOT structure. The two factors may to large degree affect the agricultural information management, including information representation, analysis and processing, information utilizing and information services. Therefore, it is very important both theoretically and in reality to strengthen the research on AIOT information management in the promotion of agricultural information intelligent management and the development of agricultural informationization.An ontology is a specification of a conceptualization, and it has unique advantages in knowledge and information representation and description, classification and relative calculations. So, the technology of ontology is adopted in this thesis to combine with such intelligent computing technologies as Neural Networks and so on to study on the intelligent capture, representation, analysis and fusion processing of AIOT information. Then the ontology-based mechanism of AIOT information intelligent management is proposed. Finally, it is exemplified with the application of IOT in the rice-duck complex agro-ecosystem. The following subjects are mainly studied and emphasized.1) Based on the analysis of the types and features of AIOT information, a novel viewpoint is proposed that AIOT information is composed of context information and relative domain knowledge and information, and that the two types of information ought to be changed into each other;2) After exploring the common means and models of AIOT information capture, an AIOT information captive and representation model is proposed. And the method of AIOT context modeling and representing is implemented in actual examples.3) In view of the features such as heterogeneity, uncertainty and the diversified representative forms of AIOT information, based on the analysis of the traditional methods of information fusion, an AIOT information fusion method based on the hybrid algorithm of ontology and neural networks and architecture is proposed in order to raise the level of information accurate management.4) After analysis and summary of new demands of agricultural information intelligent management under the environment of IOT, an ontology-based AIOT information intelligent management mechanism is proposed. And an ontology-based Demands-Functions-Applications models (ODFA) is proposed, focusing on the implementation of the seamless connection and combination of the demands of information utilizing, the function of AIOT information system and information services. Then, it is emphasized that besides technique realization, a series o f supporting management means and inspiriting mechanism should be established and formed as well. Finally, the measures and steps to promote the agricultural informationization management and establish inspiriting mechanism are proposed.5) An ontology based AIOT information utilizing model and an AIOT interactive information service model are proposed. And three AIOT information service patterns are discussed, including personalized information pushing services, self-service of information pulling, and the hybrid of the above two patterns. On the basis of the proposed models, information management mechanisms and service patterns, an AIOT information management and service platform is designed and implemented.6) It is exemplified with the application of IOT in the rice-duck complex agro-ecosystem to verify the AIOT information intelligent management. And a self-made duck feeding robot electric car is designed and implemented both in system software and hardware.The studies in this thesis cover IT and some domain knowledge in agriculture, management, etc. The achievements are promising with some theoretical and applied value.