Modern influence mines use Ship Classification Systems (SCS) to differentiate between targets and non-targets through an analysis of signature features. Unlike classical mine modelling, which follows a deterministic approach, in this paper we describe the stochastic mine model that was developed at NURC. The model, called Fuzzy Logic Advanced Mine Emulator (FLAME) is a generic mine model using a fuzzy logic approach. An important advantage offered by this model is that it can be retrained and is able to simulate modern and even future mine threats. In the experimental section, we show another advantage of the model; the capability of the FLAME to identify the signature features that offer the greatest level of discrimination for the specific target(s) of interest. Clearly, this provides valuable information for MCM forces, mine layers and ship designers. In the experimental section we show the capabilities of FLAME to learn to distinguish the signatures of two ships. Also we show how the output statistics of FLAME provides detailed insights on the most distinctive signature features of these two ships.
NATO Fuzzy Logic Generic Mine Model
COCOCCIONI, MARCO
2007-01-01
Abstract
Modern influence mines use Ship Classification Systems (SCS) to differentiate between targets and non-targets through an analysis of signature features. Unlike classical mine modelling, which follows a deterministic approach, in this paper we describe the stochastic mine model that was developed at NURC. The model, called Fuzzy Logic Advanced Mine Emulator (FLAME) is a generic mine model using a fuzzy logic approach. An important advantage offered by this model is that it can be retrained and is able to simulate modern and even future mine threats. In the experimental section, we show another advantage of the model; the capability of the FLAME to identify the signature features that offer the greatest level of discrimination for the specific target(s) of interest. Clearly, this provides valuable information for MCM forces, mine layers and ship designers. In the experimental section we show the capabilities of FLAME to learn to distinguish the signatures of two ships. Also we show how the output statistics of FLAME provides detailed insights on the most distinctive signature features of these two ships.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
