摘要
提出了利用基于邻域粗糙集的支持向量机算法实现污水处理过程中出现故障的自动诊断.首先,对污水处理厂收集的监控数据进行预处理;其次,利用邻域粗糙集模型中对象邻域的上、下近似,寻找2种类别的交界部分,从而减小问题规模;然后,通过对交界部分样本进行混淆度分析,剔除异常样本或噪声数据;同时对样本集进行属性约简与加权处理;最后,在约简集上进行支持向量机的训练与测试.将试验结果与传统支持向量机方法的结果进行比较,表明该方法不仅提高了故障诊断的效率,而且降低了问题的复杂程度,同时还保持了较好的推广性能.
In order to automatically diagnose the fault in the sewage treatment proacess, support vector machine based on neighborhood rough set was used. Firstly, data preprocessing was clone on training set from three different sides. Secondly, neighborhood rough set was used to find these samples in boundary and to obtain a reduced training set,at the same time, those abnormal samples were deleted. And then, attribute reduction was done and feature weight was imported. Finally support vector machine was trained and tested on the reduction set. The results compared with the others showed that the method not only improved the efficiency of fault diagnosis, but also reduced the complex degree and maintained a better generalization performance.
出处
《甘肃农业大学学报》
CAS
CSCD
北大核心
2013年第3期176-180,共5页
Journal of Gansu Agricultural University
基金
甘肃省科技支撑计划资助项目(1011NKCA058)
甘肃省高等学校研究生导师科研项目(0902-04
1202-04)
关键词
邻域粗糙集
支持向量机
污水处理
neighborhood rough set
support vector machine
sewage treatment
