基于粗糙集理论的系统建模与故障诊断

Modeling and fault diagnosis based on rough set theory

应用粗糙集理论,可以从原始的数据中提取有用的知识或规则。依据这一思想,本文建立了一个实际非线性系统的粗糙集模型。在建模过程中,首先用系统输入输出的采样数据构成原始信息表,然后离散化,再利用粗糙集算法得到系统粗糙集模型的不完备规则集,通过实验和线性插补法实现规则集完备化,最后完成模型的设计和校验。校验结果表明所建的粗糙集模型是有效的,并利用该模型实现了系统的一种故障诊断。

Applying rough set theory （RST）, useful knowledge or rules can be extracted from original data. According to this idea, the rough set model （RSM） of an actual nonlinear system was built. In modeling process, an original information table is firstly constructed from the sampling data of the system, then it is discretized, and the incomplete set of the RSM is obtained using RST algorithm; the completion of the rule set is realized through experiments and linear interpolation. Finally, the RSM was designed and verified. Testing results state that the built RSM is valid. In addition, the model was used to detect a fault of the system, and the detecting result is given.