Study on transformation method of numerically-controlled machine tool

Abstract

In the research of the transformation method of numerically-controlled machine tool (NC machine), due to the current methods exist the problem of that for solving process is complex and the optimization process is easy to fall into local optimal solution, a transformation method of NC machine based on improved particle swarm optimization algorithm is presented. Firstly, particle swarm algorithm and the training parameters of support vector machine (SVM) are fused, to set up the prediction model of transformation state based optimization SVM for NC machine. And then the genetic algorithm is introduced, in addition, real number coding rules and improved genetic operators are applied to optimize the design of NC machine’s spindle, to accurately complete the transformation of NC machine. Simulation results show that the accuracy of the transformation method of NC machine, based on improved particle swarm algorithm proposed in this paper is high, and practicability is strong which can meet the application requirements of the transformation of NC machine. Introduction Many China's units have introduced high-end NC machine from abroad, the service period of some of them is to the deadline, some cannot work normally, in the idle state [1], and it not only caused a great waste of resources, but also cannot meet the needs of normal production and modern high precision, high speed and high reliability requirements of processing. Therefore, the transformation of the NC machine has become the focus subject of many experts, and has been widespread concerned [2]. In literature [3], artificial visual neural algorithm is used to establish the transformation model of the NC machine, to complete the transformation of NC machine. The method has strong adaptability, but has the problems of slow convergence speed, more time-consuming problem. The literature [4] uses a transformation method of NC machine based on support vector machine algorithm. The method can effectively solve the small sample and nonlinear problem, but the solution process is complex and the optimization process is easy to fall into a local optimal solution. The literature [5] focuses on the study of the transformation of NC machine based on ant colony algorithm. Although this method is accurate, problems of easily falling into local minimum, slow convergence rate etc. are existed in it. In view of the above questions, a transformation method of NC machine based on improved particle swarm optimization algorithm is proposed in this paper. Firstly, particle swarm algorithm and the training parameters of support vector machine (SVM) are fused, to set up the prediction model of transformation state based optimization SVM for NC machine. And then the genetic algorithm is introduced, in addition, real number coding rules and improved genetic operators are applied to optimize the design of NC machine spindle, to accurately complete the transformation of NC machine. Simulation results show that the accuracy of the transformation method of NC machine, based on improved particle swarm algorithm proposed in this paper is high, and practicability is strong which can meet the application requirements of the transformation of NC machine. The transformation theory of CNC machine The principle of NC machine transformation can be detailed by the following formulas: Taking diameter d , outer diameter D and the segments span length L of spindle as design International Conference on Automation, Mechanical Control and Computational Engineering (AMCCE 2015) © 2015. The authors Published by Atlantis Press 1279 variables ( ) 1,2, , i X i n n =  is the number of design variables of the model, spindle’s optimization design variables is expressed as follows: , , i i x d D L   =   ∑ ∑ (1) ( ) 1 f x is the stiffness target function of spindle, ( ) 2 f x is the volume target function of spindle, then the optimization objective function is shown in formula (2): ( ) ( ) 1 1 1 1