论文部分内容阅读
针对生化过程故障不易发现的问题,提出了一种将支持向量机故障诊断和MCGS组态软件相结合的在线故障诊断方法。该方法借助LIBSVM工具箱在MATLAB软件中完成SVM分类器模型的训练,利用OPC接口标准实现MCGS组态软件与MATLAB软件的实时数据交换,根据训练得到的SVM分类器对实时采集到的数据进行分类判断,然后通过OPC接口将诊断结果回送给MCGS组态软件,最终在组态软件人机界面中将故障信息显示出来,从而实现故障诊断的智能化。测试运行证明了本方法的有效性。
In order to solve the problem of biochemical process failure, an online fault diagnosis method based on support vector machine fault diagnosis and MCGS configuration software is proposed. The method completes the training of SVM classifier model in MATLAB software by LIBSVM toolbox, realizes the real-time data exchange between MCGS configuration software and MATLAB software by using OPC interface standard, classifies the data collected in real time according to the trained SVM classifier Judgment, and then through the OPC interface to the diagnostic results back to the MCGS configuration software, the final configuration software man-machine interface fault information is displayed, in order to achieve fault diagnosis intelligent. Test run proved the effectiveness of this method.