中文摘要：

在利用SVM对齿轮箱进行故障诊断决策时,SVM模型参数（核参数及惩罚因子）对齿轮箱故障的诊断结果影响很大,而最优参数难以获取,针对这一问题,提出一种基于自适应细菌觅食算法（BFA）的SVM参数快速选取方法。以齿轮箱故障诊断过程为实验对象,对比分析网格搜索法、遗传算法、粒子群算法与细菌觅食算法对SVM径向基核函数参数δ及惩罚因子C的优化性能。研究结果表明：细菌觅食算法能够更加快速地选取到最优参数;采用细菌觅食算法优化SVM参数可以进一步提高齿轮箱故障诊断的精度。

英文摘要：

Using SVM to study the fault diagnosis decisions for gearboxes, the SVM model parameters（kernel parameters and penalty factor） have great influence on the gear fault diagnosis results, however, the best parameters are difficult to obtain. In order to solve this problem, a rapid selection method was proposed for SVM based on adaptive bacterial foraging algorithm（BFA）. By taking the gearbox fault diagnosis process as experimental subject, the performance of the grid search method, genetic algorithm, particle swarm optimization and bacterial foraging algorithm was analyzed to optimize the RBF kernel function＇s parameter δ and the penalty factor C. The results show that the bacterial foraging algorithm can select more quickly the optimal parameters. Using the bacterial foraging algorithm to optimize SVM can further improve the accuracy.