Research On Mechanical Process Decision Rules Mining And Process Planning Of Spacecraft Shell Parts

With the rapid development of the space project of our country,the traditional process design for typical parts of spacecraft oriented to ?multi-class,small-batch? production mode have problems of low efficiency,poor quality,and low level of intelligence.The reason are as follows: firstly,the empirical knowledge contained in the process data has not been effectively utilized;secondly,process planning still relies on engineer experience,resulting in inefficiency.Therefore,based on spacecraft shells,study on the process knowledge network,process decision rules mining,and process planning,to improve process design efficiency.In order to constitutively express the spacecraft shell parts,a structured process knowledge model is established.Based on the machining process knowledge network,the components of process knowledge and the relationship between the m is analyzed.According to the analysis of structural organization of maching features,the model of processing feature element model and processing feature relationship model are established;By defining the process step matrix and the process step forward graph,the structured expression of process step information and step sorting rules is realized;By analyzing process cost impact factors,the process evaluation rule is defined.Thus,the rationalized organization and storage forms of various process knowledge are provided.To solve the problems of difficult reuse and low reuse rate of process knowledge in the process history data of spacecraft shell parts,the method of process decision rules mining based on rough set theory is studied.Firstly,process data is pre-processed to meet decision rule mining requirements.The decision matrix of the processing method chain is established by binary coding to realize the mining of inference process decision rules.Based on the process attribute reduction,the process decision rules are mined and the process decision rule base is established.To improve the intelligent level of the process planning of spacecraft shell parts,the process planning method based on artificial specific immune system(ASIS)is studied.By analyzing the relationship between process planning and specific immune process,the process quality evaluation function is defined,and ASIS model is established.The generation of initial processes are based on the Dijkstra algorithm.And the introduction of adaptive parallel transformation operator was used to guide the dynamic adjustment of processing methods and manufacturing resources,which enabled the algorithm to have adaptive adj ustment capabilities that could follow the iterative process and changes in program quality.Finally,based on the above research,the process design tool set oriented spacecraft shell parts is developed.Feature extraction tool is developed based on Siemens NX 10.0,and tools for process decision rules mining and process planning are developed based on Matlab.Thus,the rapid modeling and feature extraction of the typical features of spacecraft shells are realized,and the purpose of processing method push and process planning is achieved.