Genetic algorithm for Lagrangian support vector machine optimization and its application in diagnostic practice

In this article a genetic algorithm optimized Lagrangian support vector machine algorithm and its application in rolling bearing fault diagnosis is introduced. As an effective global optimization method, genetic algorithm is applied to find the optimum multiplier of Lagrangian support vector machine. Synthetic numerical experiments revealed the effectiveness of this genetic algorithm optimized Lagrangian support vector machine as a classifier. Then this classifier is applied to recognize faulty bearings from normal ones. Its performance is compared with that of backpropagation neural network and standard Lagrangian support vector machine. Experimental results show that the classification ability of our classifier is higher and the computing time required to find the separating plane is relative shorter.