Research And Development Of Database Platform And Process Decision Method Of Barrel Finishing

Barrel finishing is a basic and universal manufacturing technology,which aims to improve the surface quality and integrity of parts.It has been widely used in traditional manufacturing and high-end equipment manufacturing.According to the statistics of international authoritative experts Cariapa,about 50% of the mechanical parts can be processed by barrel finishing to improve their surface quality.Due to the specialty and complexity of the barrel finishing process system,there is a dead zone of process supply and demand information among enterprises in the whole industrial chain,which seriously restricts the applications of the barrel finishing technology in the manufacturing field.Based on the study of a large number of typical manufacturing cases accumulated from practical experience and public available sources,it is an important and realistic task for enterprises in the whole industrial chain to develop a database platform for barrel finishing and to explore the process decision method of database,as well as to promote the rational sharing of effective processing information resources.It is also essential for enterprises to transform,upgrade,and improve quality and efficiency.It is significant to further broaden barrel finishing technology.The main research purposes of this project are as follows.Firstly,the construction mode of the barrel finishing database is studied,which meets the practical requirements from enterprises to directly acquire basic barrel finishing processing information.Secondly,the intelligent process decision method and application strategy of database are studied so that different users can obtain the desired solution through database according to their own needs,including the information of equipment,barrel media,barrel liquid reagents and processing parameters.First of all,the information resources for processing and the integrated model are constructed by analyzing the barrel finishing process.The development of modeling method of the database platform adopts the graphical description method,combining the integrated computer-aided manufacturing definition(IDEF)and the Unified Modeling Language(UML).The database modeling method is composed of function model,organization model,information model,knowledge model and process model.In this system,the integration relationship of the barrel finishing database with the process model as the core is established.The three-layer architecture of the barrel finishing database is established,which is composed of visual layer,method layer and application layer.It can achieve the integration of the enterprise processing environment and infrastructure in the whole industrial chain,and lay the model and system foundation for the construction of the database platform.This paper analyzes the processing cases,and combines the processing objects and requirements as the main characteristics to form the cases and assemble them into the case base,which achieves the case representation of the processing cases.The improved fuzzy C-means clustering algorithm(S-FCM)based on subtractive clustering is proposed to find and save the special cases,so as to improve the clustering quality.The other cases are compared by two similarity in order to delete the redundant cases so as to optimize the case base reasonably and effectively.A large number of simulation studies have been carried out by using the existing qualified cases of the self-developed barrel finishing database platform.The results show that the proposed method can reasonably screen and delete redundant cases in the case database to save the case storage space,which significantly improves the case retrieval efficiency and meets the real-time control of the production site.The principle of this method is simple and the steps are clear.It can be used as a database platform for the formulation of intelligent barrel finishing process and the rapid and accurate optimization of process parameters in the production process.In order to optimize the process scheme intelligently,a hierarchical fusion decision theory is proposed.According to the process case base constructed by the processing technology database,first of all,the weighted case-based reasoning technology(WCBR)is used to find the original case that matches the new problem in order to quickly find the solution of the problem.If the matching case is not found,the fuzzy expert system(FES)is used to fully dig the knowledge in the existing cases,and the similar case of the new problem is found by interval value fuzzy reasoning.Among them,a case-based reasoning method with variable weight is proposed in detail.Based on the analytic hierarchy process,the case feature weight is determined,the step of case hierarchical retrieval and the calculation method of case feature similarity are defined.The existing cases,similar cases and cases with large differences in case database are simulated and studied,and the influence of feature judgment matrix on the optimization results is discussed.The simulation results show that the matching case to the new problem can be found rapidly and accurately by using WCBR.In addition,for the cases that matching cases cannot be retrieved,a fuzzy expert reasoning model for the optimization of barrel finishing process is proposed.The optimization of barrel media is taken as an example and the construction of interval value fuzzy rules is described in detail.According to the successful cases of actual processing,the range and membership interval of each case feature value are determined,and the relationship with the parameters of barrel media is established.The production rule representation method is used.The interval value fuzzy rules are established.The weight of each characteristic attribute in the fuzzy rules is determined by the analytic hierarchy process,and the interval value fuzzy reasoning algorithm is used to design the reasoning machine for optimizing the parameters of the barrel media.A large number of test cases are used for experimental simulation,and the results show that the fuzzy expert reasoning optimization model can improve the relationship between new problems and old cases on the basis of case reasoning.At the same time,it can quickly,accurately and reasonably optimize the barrel finishing process for the parts to be processed.The overall framework and functional structure of barrel finishing database development,including physical resource layer,virtual resource layer,data management service layer,application interface layer and user layer platform,are constructed.Based on Oracle database management system,C# and Python development language,Microsoft Visual Studio 2017 integrated development environment and B/S network structure mode,the database platform program is well developed,and the practical application of the platform is shown in multiple dimensions.Based on the practical requirements of finishing database from a large aero-engine manufacturing company,the function of generating special process template is extended,and the data interface with PDM system is established.Manufacturing application shows that the processing template can facilitate standardization of processing procedure.The application of the overall database platform promotes the integration,accumulation and communication of the special process information resources among entrepreneurs,which plays a positive role in improving quality and efficiency and information management.The database platform and its process decision method developed in this paper can be directly applied to the special process decision-making and management upgrading for enterprises in the whole industrial chain.They can provide theoretical support and practical exploration for the continuous improvement of quality and efficiency of the barrel finishing industry.It also provides a useful reference for other manufacturing technologies and industrial fields to build an database platform.