Details

Autor(en) / Beteiligte

• Zhou Yu-Guo
• Zhou Xin
• Fan Yong-Le

Titel

Nonlinear filtering application in fault diagnosis

Ist Teil von

• 2017 Chinese Automation Congress (CAC), 2017, p.2789-2792

Ort / Verlag

IEEE

Erscheinungsjahr

2017

Quelle

IEEE Xplore

Beschreibungen/Notizen

• The nonlinearity estimation is a major problem in the application of Kalman filtering. Though EKF algorithm achieves the output estimation by using approximate nonlinear function, and it enhances the application in nonlinear system to some extent, for strong nonlinear systems, the EKF still has a large estimation error. Therefore, an approximate probability density of the UKF algorithm is proposed. The approximate probability density function is used to replace the traditional approximate nonlinear function. The simulation results show that the proposed method is feasible.