Modular fault diagnosis for ROVs based on a multi-model approach

The publication on hand presents the results of the ongoing developments of a modular fault detection algorithm for ROVs within the MiroBase Project. It is based on a multi-model approach that was developed, applied and tested in a MATLAB/Simulink environment. It is planned to test the implementation on a specific ROV under real conditions. The MiroBase project aims to provide a framework for modular ROVs. The modularity will allow different hardware and software configurations. In order to enable Fault Detection and Isolation (FDI) for these different configurations it is necessary to develop an algorithm that can be adapted quickly and easily. A further challenge is the possibly incomplete sensor equipment, which prevents complete monitoring of all components. This paper describes the following steps that are necessary to implement model-based fault diagnosis for ROVs: designing the model, generating residuals, calculating fault models, determining fault type and quantity.