FIELD: hydro acoustics.
SUBSTANCE: invention relates to underwater acoustics and can be used to implement neural network recognition of target classes (surface or underwater object) detected by the signs of amplitude-phase modulation of low-frequency signals pumping the marine environment by radiation and fields of objects. Method of detection and neural network recognition of signs of fields of various physical nature, generated by marine targets, is that in the marine environment they form a zone of non-linear interaction and parametric conversion of pump waves with object signals, for which the radiator and the receiving antenna are placed at opposite boundaries of the monitored section of the marine environment. Pump waves, modulated by object signals, receive and amplify in a parametric conversion band, transfer their time-frequency scale to the high-frequency region, conduct narrowband spectral analysis, emit parametric components of the total or difference frequency, which, taking into account the time and parametric transformation of the waves, restore the characteristics of the object signals. Principal difference from the prototype is that the amplitude-frequency characteristics of the object signals, obtained using narrowband spectral analysis in the path of receiving, processing and recording signals, are fed to the additionally introduced path of neural network recognition and classification. Path of neural network recognition and classification consists of a target class recognition unit based on amplitude-frequency characteristics that implements the computational operations of an artificial neural network and is covered by feedback from the learning unit, in whose memory the data of mathematically processed spectrograms of marine targets are recorded. At the first input of the target class recognition unit, the amplitude-frequency characteristics receive data from the output of the spectrum analyzer of the receiving, processing and recording signal path, and the second input receives data from the neural network recognition and classification path learning unit. Artificial neural network is configured according to the classification criteria of fields generated by marine targets, they start computing operations, by the results of which they correct its weights and form a conclusion about the degree of belonging of the studied spectral region to the object of classification (surface or underwater object). This technical result is achieved by applying the computational operations of an artificial neural network using an operatively updated library of mathematically processed spectrogram images of marine targets.
EFFECT: provision of neural network recognition of classes of marine targets (surface or underwater object), detected by the signs of amplitude-phase modulation of low-frequency signals of pumping of the marine environment by radiation and fields generated by objects.
1 cl, 5 dwg
