论文部分内容阅读
机械故障诊断是典型的模式识别问题,遗传编程能够较容易地构造大规模特征集;可以根据分类效果自动进行特征选择,从而获得优质的最佳特征。通过滚动轴承分类器的设计验证了算法的准确性。
Mechanical fault diagnosis is a typical pattern recognition problem. Genetic programming makes it easier to construct large scale feature sets. Feature selection can be performed automatically based on the classification results to obtain the best features of high quality. The design of the rolling bearing classifier verifies the accuracy of the algorithm.