Underwater signal interpretation in home system

In recent years, the application of artificial intelligence technique to underwater acoustic environment obtains great attention. This paper presents the work on developing a Micro-Expert System for underwater target classification. The major contributions include that: 1) A new tri-symbol coding method for symbolic representation of waveforms is proposed, and successfully applied to knowledge-based waveform analysis; 2) The properties, features and deployment conditions of a variety of underwater ECM (electric counter-measure) equipments together with target features are collected and summarized and a combined knowledge representation approach is developed for application in underwater acoustic environment; 3) A suitable Blackboard model inference mechanism is developed to cope with time sequence and uncertainty information and 4) A Micro-Expert System Shell for underwater target classification is developed through integrating the above special techniques. The current system shell is flexible, friendly with user-interface, and easy to be extended. Some experimental results are given to show the effectiveness of this system shell.