Method for detecting fault of train suspension system on basis of robust observer

The invention relates to a method for detecting a fault of a train suspension system on the basis of a robust observer. The method comprises the following steps of: a, carrying out modeling on the train suspension system and designing the corresponding fault detection robust observer; b, utilizing an acceleration sensor and a gyroscope to respectively acquire acceleration signals and angular speed signals of a vehicle body and front and rear bogies when a train operates; c, carrying out treatment of anti-aliasing filtering, high-pass filtering, quadratic integration and the like on the acceleration signals and only carrying out anti-aliasing filtering and high-pass filtering on the angular speed signals to obtain a system output; d, utilizing the fault detection robust observer to carry out processing on the system output to obtain a residual output; e, according to a residual sequence, applying a MCUSUM (Multivariable Cumulative Sum) method to calculate a fault detection index; and f, by a fault diagnosis expert system, generating a fault alarm threshold value in real time and judging whether to give an alarm. According to the detection method, when the suspension system has a small fault, the fault can be detected in real time, the alarm is given and the possibility of false alarm and failure in report of a fault detection system can also be reduced.