International Journal of Neural Systems Engineering
Volume 1, No. 1, 2017, pp 7-14 | ||
Abstract |
An improved SVM model with double quantum genetic optimization and its application in fault diagnosis of rolling bearing
|
Support vector machine (SVM) is a kind of intelligent pattern recognition method which is suitable for small samples, nonlinear, high dimension and other complex problems, But its classification performance and generalization ability are limited by the choice of penalty coefficient and kernel function parameters. To solve this problem, in this paper, a new SVM model optimized by quantum genetic algorithm is proposed. The objective function is the classification accuracy, and the parameters of penalty function and kernel function are optimized to improve the local search ability and parameter optimization of classical genetic algorithm. A new quantum gate algorithm using a dual quantum genetic algorithm improves quantum mutation operations and catastrophic strategies. The experimental results show that the improved SVM model has higher classification accuracy, higher recognition rate and convergence rate than the traditional algorithm, and is of great significance to the development of fault diagnosis of rolling bearing.