论文部分内容阅读
本文在最小二乘支持向量机(LS-SVM)框架,针对Hammerstein非线性系统,设计了一种数据驱动故障诊断与分离(FDI)方法。具体内容包括:构造单适当的输出系统设计基于等价空间的残差生成器,进一步将残差生成器扩展至多输出情况。为了解决实时监控问题,通过低阶和稀疏逼近理论构造半参数残差生成器,此残差生成器带有参数动力学方程和非参数静态函数。在半参数框架下,通过设计参数化矩阵重构实现对传感器/执行器故障的检测与分离。最后,本文给出一个仿真实例验证了所提方法的有效性和实用性。
In LS-SVM framework, this paper designs a data-driven fault diagnosis and separation (FDI) method for Hammerstein nonlinear systems. The concrete contents include: Constructing a proper output system The residual generator based on equivalent space is designed, and the residual generator is further extended to multiple output cases. In order to solve the problem of real-time monitoring, a semi-parametric residual generator is constructed by using the theory of low-order and sparse approximation. The residual generator has parametric dynamic equations and non-parametric static functions. Under the semi-parametric framework, sensor / actuator fault detection and separation are realized through the design of parametric matrix reconfiguration. Finally, a simulation example is given to verify the effectiveness and practicability of the proposed method.