Abstract: In order to effectively diagnose the fault of the equipment，a support vector machine algorithm based on parameter optimization was proposed.Firstly，the immune clonal selection mechanism was introduced.The antibody group was constructed，in which an antibody was expressed in two decimal numbers.Then，the affinity function was established on the basis of the missing report rate.Thus，the parameters of support vector machine were optimized.After that，the support vector classifier with optimized parameters was constructed and used to classify and detect the equipment data.Simulation verification was carried out on the data of turbo generator set.The simulation results indicate that the proposed algorithm does not significantly increase the training time，compared with the traditional support vector machine algorithm.Moreover，it can effectively improve the detection rate and reduce the false detection rate.

Key words: electromechanical engineering, fault diagnostics, support vector machines, parameter optimization, immune clone algorithm

摘要： 为了有效的诊断出设备的故障，给出了一种基于参数优化的支撑矢量机算法。该算法首先引入免疫克隆选择机制，以两个十进制数表示一个抗体来 构建抗体群，以漏报率为基础构造亲和度函数，实现支撑矢量机参数的优化。然后使用优化后的参数构造支撑矢量分类器对设备数据进行分类检测。通过 在汽轮发电机组的数据集上进行仿真验证，实验结果表明，该算法相对传统的支撑矢量机算法不会显著增加训练时间，并且能够有效提高检测率和降低误 检率。

关键词: 机电工程, 故障诊断, 支撑矢量机, 参数优化, 免疫克隆算法