Research On The Method Of The Intelligent Feature Recognition And Nonlinear Process Planning For STEP-NC 2.5D Milling Manufacturing Feature

In recent years,along with the rapid development of advanced manufacturing technology and the ever-increasing demands for processing quality and processing efficiency,numerical control systems based on the traditional ISO 6983 standard increasingly expose their defects in structure,function and human-computer interaction,To a certain extent,restricted the progress of numerical control technology and the development of productivity.To this end,a new data interface STEP-NC for modern computer numerical control system has been proposed in the world,which provides a common product data exchange model for CNC machining.It covers all the information of product processing,which makes the machine complete the adjustment and optimization of the product processing under the condition of "understanding" the product provides a new platform for the development of the CNC system in the direction of intelligence,integration,flexibility and networking.In this paper,STEP-NC 2.5D milling manufacturing features as the research object,its intelligent feature recognition and nonlinear process planning methods are studied,the main contents are as follows:（1）Introduced the origin and advantages of STEP-NC standard,described the status quo of STEP-NC,STEP-NC feature recognition and process planning at home and abroad,analyzed the existing problems in various research methods,and on this basis put forward the research direction and train of thought of this article.（2）Discussed the architecture of the STEP standard,the components,and the EXPRESS language used to describe the method in this standard.The STEP-NC standard and its data model are introduced,and the ISO 14649 standard is selected as the research standard of this paper through the detailed explanation and comparison of the two standards of STEP-NC.（3）Taking the feature of STEP-NC 2.5D milling manufacturing as the research object,a feature recognition method based on the minimum sub-graph and the improved BP neural network is proposed.The method started from the STEP AP203 file,preprocessed its geometric topology information,constructed the minimum subgraph based on the extracted face and edge information and the concave and convex properties of the edge,and generated the input vector of BP neural network,And then used chaos genetic algorithm to optimize BP neural network to get the output vector of BP neural network through training to get the result of feature recognition.（4）A nonlinear process planning method for manufacturing features of STEP-NC 2.5D milling was proposed.The method included three parts:macroscopic,detailed and microscopic process planning.In the macro process planning stage,the BP neural network based on chaos genetic algorithm was used to determine the machining operation method of each feature.In the detailed process planning stage,In the microscopic process planning stage,the machining resources were matched according to the rules,and chaos genetic algorithm was used to optimize the sequence of work steps and process parameters to obtain the result of nonlinear process planning.（5）With Visual C++6.0 as the development environment,using MFC application development framework,the intelligent feature recognition and nonlinear process planning prototyping system oriented to STEP-NC milling manufacturing features under Win7 system platform were developed.