A Distributed Diagnosis Strategy using Bayesian Network for Complex Wiring Networks

In this paper, a distributed diagnosis strategy using reflectometry is proposed. It consists in making reflectometry measurements at different spots of a highly complex wiring network simultaneously in order to reduce ambiguities caused by multipath signals propagation. Although the problem of sensors number optimization is greatly studied in the literature, it is not well investigated in complex wiring networks diagnosis. The proposed approach is based on two principles which are sensors number and location optimisation using Bayesian Networks and measure uncertainty estimation. It consists in four steps: (1) sensors implementation in a deterministic case, (2) influential parameters on diagnosis measure identification, (3) diagnosis measure modelling using Bayesian Networks, (4) sensor number and location optimization. Here, the main objective is to minimize both sensors number and diagnosis measure uncertainty.