Applying genetic algorithms on fuzzy logic system for underwater acoustic signal recognition

The propagation of underwater sound is affected due to multi-path, reverberation and inhomogeneity. In addition, the random process and time varying characteristics of the underwater signal lead to the difficulty in using an effective mathematics model to simulate the system. Hence, a structure of fuzzy logic system optimized by genetic algorithms was developed for the task of recognition of underwater acoustic signal in real time. The fuzzy logic algorithm was used for defining the linguistic variables of the feature. Based on these variables the membership function of the fuzzy rule was defined. In order to improve the recognition rate, a genetic algorithms was developed in the proposed system frame; this genetic algorithm was used in system training phase for the purpose of adapting the membership functions of the inference rules. The simulation results have demonstrated the effective performance of the proposed system.