A Comparison Of Inverse Simulation-Based Fault Detection In A Simple Robotic Rover With A Traditional Model-Based Method

Abstract

Robotic rovers which are designed to work in<br> extra-terrestrial environments present a unique challenge in terms<br> of the reliability and availability of systems throughout the mission.<br> Should some fault occur, with the nearest human potentially millions<br> of kilometres away, detection and identification of the fault must<br> be performed solely by the robot and its subsystems. Faults in<br> the system sensors are relatively straightforward to detect, through<br> the residuals produced by comparison of the system output with<br> that of a simple model. However, faults in the input, that is, the<br> actuators of the system, are harder to detect. A step change in<br> the input signal, caused potentially by the loss of an actuator,<br> can propagate through the system, resulting in complex residuals<br> in multiple outputs. These residuals can be difficult to isolate or<br> distinguish from residuals caused by environmental disturbances.<br> While a more complex fault detection method or additional sensors<br> could be used to solve these issues, an alternative is presented here.<br> Using inverse simulation (InvSim), the inputs and outputs of the<br> mathematical model of the rover system are reversed. Thus, for a<br> desired trajectory, the corresponding actuator inputs are obtained.<br> A step fault near the input then manifests itself as a step change<br> in the residual between the system inputs and the input trajectory<br> obtained through inverse simulation. This approach avoids the need<br> for additional hardware on a mass- and power-critical system such<br> as the rover. The InvSim fault detection method is applied to a<br> simple four-wheeled rover in simulation. Additive system faults and<br> an external disturbance force and are applied to the vehicle in turn,<br> such that the dynamic response and sensor output of the rover<br> are impacted. Basic model-based fault detection is then employed<br> to provide output residuals which may be analysed to provide<br> information on the fault/disturbance. InvSim-based fault detection<br> is then employed, similarly providing input residuals which provide<br> further information on the fault/disturbance. The input residuals are<br> shown to provide clearer information on the location and magnitude<br> of an input fault than the output residuals. Additionally, they can<br> allow faults to be more clearly discriminated from environmental<br> disturbances.