Research On Intelligent Numerical Control Programming Technology Of Key Parts Of Marine Diesel Engine Based On Knowledge Graph

For a considerable period,the manufacturing process of marine diesel engines has lagged behind the overall ship manufacturing,constituting a significant bottleneck which limits the progress of the shipbuilding industry.As a crucial constituent of marine diesel engines,the key parts of marine diesel engines possess intricate numerical control processes,rich parts,and varied processing characteristics,which leading to issues of low efficiency,high labor intensity,and excessive dependence on experience in CAM NC programming.Even though marine diesel engine enterprises have amassed a vast collection of technical documents for CAM NC programming,forming a relatively comprehensive knowledge structure,CAM NC programming efficiency remains insufficient in practical performance.The explanation for this inadequacy is that numerous NC process documents,historical programming cases,and NC programming experience in companies have not been properly represented,stored,and managed,and new programming tasks cannot be efficiently associated with existing knowledge,making knowledge reuse difficult to achieve.Therefore,to enhance the representation,management,storage,and reuse capabilities in the field of CAM NC programming and subsequently raise the automation level and intelligence degree of NC programming of key parts of marine diesel engines,this thesis focuses on the construction and application of knowledge graph-related technology as the core.The main research objectives are outlined below:(1)Research on knowledge graph construction technology for CAM NC programming.Firstly,in order to accurately describe the information in the field of CAM programming,CAM information and feature-operation information model was built.the CAM processing operation template library with typical characteristics was extracted from the CAM programming cases of enterprise history.A top-down knowledge graph construction mode was adopted and a knowledge graph construction method for CAM numerical control programming was adopted.Taking the processing operation template library as the data source,the CAM knowledge graph data layer was constructed by means of information extraction,knowledge extraction and knowledge fusion.(2)Research on NC programming technology driven by instance knowledge graph.The matching model between features and processing operation templates was constructed by GA-BP algorithm,which improves the rationality of CAM template selection.Through the knowledge extraction of features and templates,the feature instance knowledge graph was constructed,and the graph was evaluated and optimized according to the actual needs and programming conditions.Aiming at the empirical knowledge in the instance graph,the method of knowledge reasoning based on ontology semantics and SWRL rules was adopted to complete the evaluation and update of empirical knowledge.Aiming at the numerical cutting parameters of the instance knowledge graph,the SVM-NSGA-II algorithm was used to optimize the multi-objective optimization with the processing time and carbon emissions per unit milling volume as the optimization objectives,and the optimization of cutting parameters was realized.The optimized instance knowledge graph was reflected as a processing operation template,and the adaptive processing CAM template parameters was realized.(3)Intelligent CAM NC programming system development for key parts of marine diesel engines.With the above technical research results as the theoretical basis,the intelligent CAM NC programming system of key parts of marine diesel engines was developed with reference to the needs of enterprises,and the modules of process information acquisition,knowledge graph decision making and NC programming were developed,and the feasibility of the system was verified by examples.