The Research On Deep Roundness Error Prediction Based Oh T-S Fuzzy Support Vector Machine Method

Deep-hole drilling of efficient precision,low cost and environmentally friendly is urgently needed proposed by aerospace,new energy equipment manufacturing and high-tech industries.However,due to the complexity of the mechanism of deep hole drilling,and how to ensure the process under normal operating conditions and precisely-real-time control the dynamic behavior of the tool,and effectively predict the quality of the machined hole,furthermore workpiece of satisfying the predetermined processing requirements by deep-hole drilling has become a hot issue of deep drilling research.Using D-optimal experimental design and combined with quadratic relationship model by response surface method,the relationship between cutting parameters and machining roundness error is preliminarily verified.In addition,the variation of roundness error by deep-hole drilling is qualitative analyzed with the change of the spindle speed,the feed rate and vibration suppres-sion device's current strength.In order to enhance dynamic performance and generalization ability of identification of deep-hole drilling roundness error,fuzzy input space partitioning problem of tool vibration mode is converted into initial value problem of the initial input space.And combined with dynamic adaptive fuzzy model of coexistence between recursive links and support vector machine,a new T-S fuzzy support vector machine algorithm is proposed.Combined with the actual experimental of deep-hole drilling,how the comprehensive function of deep-hole drilling process parameters(spindle speed and the feed rate)and vibration suppression device parameters(current strength and position of vibration suppression device)affects tool vibration and machining hole roundness error is researched.Compared the experimental results with theoretical predictions,the proposed T-S fuzzy support vector machine algorithm is verified to be effective and feasible.This method provides the technical support for online monitoring of machining hole's quality in deep-hole drilling process and active suppression of tool vibration.