Ontology-based Research On Integrated Diagnostics Of The Diesel Engine

Thanks to development of science and technology, weaponry's reliability and security problem increasingly attracts military's close attention. In order to improve the fault diagnosis and maintenance capability of the high-tech equipments, it is necessary to integrate all factors posing the equipments' testability and diagnostic capability, optimize the allocation of various diagnostic resources and study the integrated diagnostic techniques so that the detection and diagnosis ability of overall system can achieve the optimization.The integrated diagnostic techniques regard the equipments' life-cycle process as an organic whole. Therefore, to establish a unified diagnostic information model and based on this to realize the integration of diagnostic resources and the effective communication of diagnostic information is the key to achieve equipments' integrated diagnosis. Ontology can describe various types of testing and diagnostic information and their mutual relations in a standard way, so it is suitable for the establishment of a unified information model to meet the needs of equipments' integrated diagnosis.This thesis is funded by the "integrated diagnostic techniques of weaponry's electromechanical systems" which is a "11th Five-Year Plan" national defense advanced research project. The paper took certain type of ship diesel initiative device as research object, analyzed the knowledge structure in the field of equipments' integrated diagnosis by ontology, built an ontology model of integrated diagnosis which includes three sub-ontology models based on product domain, test domain and diagnosis domain. And the paper carried out in-depth studies on the model's knowledge storage and update, model building, mapping and integration methods, the design method of fusion and decision-making system based on ontology model was given. The main research accomplishments and conclusions of the thesis are as follows,1. The major problems in equipments' fault diagnosis and maintenance decision-making have been analyzed. Based on the summary of basic elements and techniques of integrated diagnosis, the merits of the application of ontology on the standardized expression of multi-source and heterogeneous diagnostic information and the problems need to be solved have been discussed.2. The paper expatiated the theoretical foundation of ontology and analyzed the general issues of ontological modeling and linguistic description. Based on ontology and needs and problems existed in the field of integrated diagnosis, the modeling principle and method according to the features of knowledge-based object were given and application of OWL into the development of integrated diagnosis ontology has been analyzed.3. Targeting on the difficulties in knowledge expression in integrated diagnosis, the knowledge-oriented ontology has been introduced. The paper established a integrated diagnosis knowledge ontology system which contains three sub-ontology models based on product domain, test domain and diagnosis domain and proposed a method by using XML to store and retrieve knowledge into and from database. An effective knowledge representation method was given for ontology modeling in the field of equipments' integrated diagnosis.4. For the main the problem and difficulties exsited in the construction of the ontology model used in the equipment intergrated diagnostic domain, the paper provided the product domain ontology construction methods based on the proposed bond graph model, the test domain ontology construction methods based on TOP-DOWN and the diagnosis domain ontology construction methods based on the fault tree. It also provided the knowledge integration method based on the Ontology Model mapping, and provided system building principles for establishing the consistent diagnostic information model.5. The paper provided an implement construction based on the diagnostic ontology model. It also provided a development and implement method ontology-oriented diagnostic decision-making system and a practical application by the Protege software on the background of a certain type of ship diesel initiative device fault diagnosis and maintenance.