MODEL-BASED SENSOR SUPERVISION IN INLAND NAVIGATION NETWORKS: CUINCHY-FONTINETTES CASE STUDY

Abstract

In recent years, inland navigation networks benefit from the innovation of the instrumentation and SCADA systems. These data acquisition and control systems lead to the improvement of the management of these networks. Moreover, they allow the implementation of more accurate automatic control to guarantee the navigation requirements. However, sensors and actuators are subject to faults due to the strong effects of the environment, aging, etc. Thus, before implementing automatic control strategies that rely on the fault-free mode, it is necessary to design a fault diagnosis scheme. This fault diagnosis scheme has to detect and isolate possible faults in the system to guarantee fault-free data and the efficiency of the automatic control algorithms. Moreover, the proposed supervision scheme could predict future incipient faults that are necessary to perform predictive maintenance of the equipment. In this paper, a general architecture of sensor fault detection and isolation using model-based approaches will be proposed for inland navigation networks. It will be particularized for the Cuinchy-Fontinettes reach located in the north of France. The preliminary results show the effectiveness of the proposed fault diagnosis methodologies using a realistic simulator and fault scenarios.