摘要
支持向量机是柴油机进行故障诊断的重要工具,然而其核函数参数和惩罚因子的不同取值会影响到分类结果正确率。针对这一问题,提出利用改进人工蜂群算法对支持向量机相关参数进行选择优化,在该方法中,将参数作为食物源,而分类正确率作为适应度函数。接着,通过对1个UCI数据集进行分类测试,证明了该改进方法的优越性:既继承了传统方法的优点又减少了收敛时间。最后将其用于柴油机故障诊断实验,进一步证明了该方法不仅能够获得较高的故障分类正确率,而且与传统人工蜂群算法相比,能够有效降低运行时间。
Support vector machine is an important tool for diesel engine fault diagnosis, but the different values of its kernel function parameters and penalty factor will affect the classification accuracy. Thinking of the problem, this paper uses improved artificial bee colony to optimize parameters of support vector machine, and in the method, parameters will be as food sources, the classification accuracy as the fitness function. Through one UCI datasets test, it demonstrates the superiority of the improved method: inheriting the advantages of the traditional method and reducing the convergence time. Finally, the method is for diesel engine fault diagnosis, and the results prove that this method not only obtains higher fault classification accuracy, but also compared with the traditional artificial bee colony algorithm, it effectively reduces the running time.
出处
《智慧工厂》
2017年第6期81-86,共6页
Smart Factory
关键词
改进人工蜂群算法
支持向量机
参数选择
柴油机故障诊断
Improved artificial bee colony
Support vector machine
Parameter optimization
Diesel engine fault diagnosis