Evaluation of Cylindricity Error Based on an Improved GA with Uniform Initial Population

Cylindrical feature has become one of the most important features in mechanical designs. This paper presents the mathematical definition of cylindricity tolerance and mathematical model for cylindricity error evaluation. An intelligent evaluation method is provided for the evaluation of cylindricity error in this paper and the innovation can be stated as follows. Firstly, instead of considering partial components of the directional vector only, this algorithm proposed takes all the components of directional vector of the axis of cylinder into consideration. Secondly, for making the genetic algorithm converge fast, an improved genetic algorithm with uniform initial population was proposed to find the optimization of the error of cylindricity. The proposed algorithm is efficient and easy to code, and the experiment results indicate its effectiveness.