Research Of Maritime Scarch And Rescue Decision Support Based On CBR

In this day and age, maritime safety becomes more and more important. Maritime search and rescue, an essential part of maritime safety, is attracting growing attention. That how to improve the level and success rates of maritime search and rescue with computer assistant decision-making tool has become the main aspect of present maritime search and rescue research.By comparison with traditional rule-based reasoning, CBR(Case-Based Reasoning) has the advantages of simplified knowledge acquisition process, dispense with complete theory model and so on. It has been widely used in the fields of decision supporting, fault diagnosis and other fields. Therefore, this paper did research on the application of CBR to maritime search and rescue decision support.This paper firstly relied on the maritime search and rescue decision support system based on multi-agent and information fusion, a scientific research project of department of education of Liaoning province. Later, by analyzing maritime search and rescue decision-making process and combining rescue center survey with present condition of maritime search and rescue assistant decision system, it came up with a CBR-based maritime search and rescue decision support model. Meanwhile, there was an implementation of the support system with multi-agent technique on the JADE platform in the end of this paper. On the application of CBR in the field of maritime search and rescue decision support, this paper conducted the following research.In theory, this paper firstly analyzed the CBR reasoning flow, focusing on the key techniques used in the reasoning process. Since it's necessary to achieve the self-study mechanism in CBR, this paper attempted to realize this system with multi-agent. Consequently, it goes further with an analysis of the fundamentals of multi-agent technology.In practice, this paper designed a CBR-based maritime search and rescue decision support model, which is based on a profound research into the decision-making process of maritime search and rescue and CBR reasoning flow. With an analysis of maritime search and rescue cases, a design of knowledge representation method of maritime search and rescue cases was putted forward along with a retrieving algorithm design of maritime search and rescue cases by combining traditional Nearest-Neighbor algorithm, Euclidian distance computing formula and structure similarity. Besides, there were requirements analysis and modules design of the system based on CBR, also a maritime search and rescue case base design. A brief implementation of key functions of this maritime search and rescue decision support system based on CBR and multi-agent was presented on JADE in the end of this paper.