Abstract: In order to improve the performance of diesel engine fault diagnosis, a method of multi core SVM fault diagnosis for diesel engine based on dimension measurement is proposed. Firstly, the internal combustion engine dimension measurement is defined, and the dimension measurement system and process design are carried out; Secondly, the multiple kernel SVM characteristics, the input fault features from the original data space is mapped to a high dimensional data space, and the use of Lagrange method to achieve primal dual solution, and then follow the principle of Mercer realize diesel multiple kernel SVM fault diagnosis; Finally, the number of kernel functions and the recognition rate of different kernel functions are given, and the advantages of the proposed method are verified by comparing the calculation time and accuracy.

Key words: Dimension measurement, internal combustion engine; multi core SVM, fault diagnosis, Mercer principle

摘要： 为提高柴油内燃机故障诊断效果，提出基于维度测定的柴油内燃机多核SVM故障诊断方法。首先，对内燃机维度测定进行定义，并进行维度测定系统和流程设计；其次，利用多核SVM特性，将输入故障特征由原始数据空间映射至高维数据空间，并利用Lagrange方法实现原始问题对偶化求解，然后按照Mercer原理实现柴油内燃机多核SVM故障诊断；最后，通过实验测试给出选取不同核函数情况下的所需核函数数量及识别率情况，并通过对比实验验证了所提方法在计算时间和识别精度上的优势。

关键词: 维度测定, 内燃机, 多核SVM, 故障诊断, Mercer原理