中文摘要：

针对最小二乘支持向量机模型中存在的核函数选择和参数确定问题，提出了基于改进蚁群算法优化改进最小二乘支持向量机的柱塞泵故障诊断模型。该模型采用径向基核函数与多项式核函数的组合作为改进LSSVM模型的核函数，并通过改进的蚁群算法搜索模型的最优参数。仿真验证表明：该模型在训练时间和诊断准确度两个指标上有较大优势，是一种有效可行的故障诊断方法，具有良好的应用效果。

英文摘要：

For the problems of how to choose the kernel function and how to determine the parameters in the LSSVM model, this paper proposed IACO-ILSSVM faults classification model on the fault diagnosis of plunger pumps. This model improved the kernel function of LSSVM model that combines the polynomial and the RBF kernel functions, and searched the optimized model parameters using improved ant colony algorithm. The simu- lation results showed that the model had great advantages on training time and diagnostic accuracy, and it was a practical fault diagnosis method with good application prospects.