An improved particle filter algorithm for mobile robot fault diagnosis with incomplete models

An improved particle filter was presented to diagnose mobile robots with incomplete models. Two features were obtained from sample-based expression for a posteriori probability density: sum of unnormalized weights of particles, and maximal a posteriori probability (MAP). They were employed to decide whether the estimation given by particle filter was believable or not. Generally, when these features were less then some pre-defined thresholds, the estimated state was unauthentic and the true state was thought to be 'unknown fault'. The thresholds were obtained off-line. This method preserves the merits of particle filter and can diagnose known faults as well as detect unknown faults. The method is testified on a mobile robot.