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 ...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.展开更多
基金The project is supported by the National Science Funds and the High Education Funds for Science andTechnology
文摘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.
