Optimization Of Blade Processing Parameters Based On Data Mining

With the continuous development of the aviation manufacturing industry,higher and higher performance requirements are put forward for aero engine blades,and higher and higher precision requirements are also put forward for the processing of titanium alloy fan blades.A large number of studies have shown that in the actual blade processing,setting the blade processing parameters has a great influence on the blade contour error.Therefore,optimizing the processing parameters of titanium alloy blades is one of the important means to reduce the error of blade contour.In this thesis,aiming at the problem of out-of-tolerance in blade processing of China Academy of Aviation Manufacturing Technology,blade processing parameters are optimized based on statistical analysis and decision tree classification in data mining technology to reduce the blade contour of titanium alloy blades.This has certain guiding significance.First,according to the cross-sectional measurement data of titanium alloy fan blade contour,the blade contour error analysis algorithm is proposed.Based on the principle of minimum area,a matching algorithm for actual measurement points and theoretical cross-section lines is proposed,a matching objective function is established,and a method for matching actual measurement points and theoretical cross-section lines is given.The calculation method of the contour error of the blade profile is established,and the contour meeting the error requirements is considered as the optimization target of data mining processing parameters.Secondly,a reverse calculation and extraction method of blade processing parameters is proposed,and the key processing parameters that affect blade profile contour are analyzed.Taking titanium alloy fan blades as the research object,a reverse calculation algorithm of tool position data based on CNC machining program is given,and the calculation and extraction methods of processing parameters such as tool attitude angle are proposed.Using multivariate statistical techniques,it is analyzed that the tool size,cutting depth,feed speed and tool tilt angle are the key processing parameters that affect the contour error of the blade profile,and these parameters are used as data mining optimization variables.Thirdly,a blade processing parameters optimization method based on decision tree is proposed.The advantages of decision tree applied to optimization of processing parameters are analyzed.The key processing parameters and the blade profile contour data process decision data table is constructed,and the information gain rate is used to measure the importance of processing parameters as the split attribute of the decision tree.The blade processing parameter optimization decision tree model was established.The blade processing parameters optimization rules whose optimization target is the blade profile contour are obtained.Finally,the actual blade processing experiment is carried out to verify the data mining-based blade processing parameters optimization method proposed in this paper,and the profile contour analysis of the shaped blade verifies the feasibility of the method proposed in this paper,in order to reduce the blade processing contour error Technical support is provided for improving blade processing accuracy.