基于C-BRB模型的发动机运行状态监测方法

针对发动机运行状态监测过程中发动机内部多个因素之间相关性与建模方法可解释性问题,提出数据驱动下C-BRB方法。该方法首先通过样本数据计算发动机内部多个因素之间的Kendall秩相关系数,并确定具体Copula模型及参数λ,实现对多个因素之间相关性的测量;然后使用置信规则库(BRB)对发动机内部多个因素建模,在BRB推理过程中,每条激活规则的综合匹配度由Copula模型对该规则中各前提属性的匹配度进行计算得到,并利用证据推理(ER)算法对所有激活规则进行融合得到输出。实例结果表明,所提方法在推理发动机传感器数据上具有较高的精度。

Monitoring Method of Engine Operation State Based on C-BRB Model

Aiming at the correlation between multiple factors inside the engine and the interpretability of the modeling method during the monitoring of engine operation state, a data-driven Copula-belief rule base(C-BRB) method is proposed in this paper. Firstly, the Kendall rank correlation coefficients between multiple factors inside the engine are calculated through the sample data, the specific Copula model and parameter λ are determined to measure the correlation between multiple factors. Then, the belief rule base(BRB) is used to model multiple factors inside the engine. In the BRB reasoning process, the comprehensive matching degree of each activated rule is calculated based on the matching degree of the attributes in the activated rule using the Copula model, and the evidence reasoning(ER) algorithm is used to fuse all activated rules to obtain the output.The results of the example show that the method proposed in this paper has high precision in reasoning the data of engine sensor.