Research On The Knowledge Base Construction And Reasoning Mechanism Of Salvage Expert System

Maritime search and rescue (SAR) works play a vital role to save lives and the loss of huge amount of property at maritime accidents. The probability of success (POS) for a SAR work, to some extent, depends on the time of emergency response and the speed of implementation of rescue work. Therefore, it's necessary to research on the technology of fast approaching to the targets in distress at sea. Based on the theory and technology of expert system, research on the knowledge base construction and reasoning mechanism of Salvage Expert System is a sub-task of a national science and technology supporting project named "the technology of fast approaching to targets in distress at sea". The key of this sub-task is building salvage expert knowledge and study on reasonable reasoning mechanism. It provides effective support for fast approaching to distress ships, makes rescue operations faster, more accurate and effective. Main tasks are as follows:Firstly, by investigating and consulting a large number of books or other information accesses on maritime SAR, understand marine casualty assistance process, acquire knowledge such as:the shore SAR site, dedicated salvage ships and helicopters and other SAR forces, marine meteorology, ocean currents, waves and other meteorological and hydrological knowledge, the emergency rescue plans and operational procedures of various dangerous situations.Secondly, a salvage expert knowledge base is established. Study on the knowledge and data for SAR, and build the knowledge model based on their characteristics.In this knowledge base, production rule representation is used to represent SAR knowledge, relational database tables is used on the storage of rules rationally and efficiently and, primary and foreign keys is used to interrelate the rules and represent the relationship between knowledge.Finally, study on the system reasoning mechanism. According to knowledge representation of salvage expert knowledge base and "International Aeronautical and Maritime Search and Rescue Manual" which was jointly launched by the International Civil Aviation Organization and IMO made rescue reasoning knowledge. With appropriate use of the knowledge reasoning model combined forward reasoning strategy, substitution of rules and the special case, gives rescue decision-making proposal and guides works of rescue.Applying expert system technology to maritime emergency SAR, can meet particular requirement for timeliness of quickly and scientific decision-making to emergencies, change the pattern of emergency rescue command relied on manual analysis, judging and decision-making, thereby increasing the efficiency and success rate of emergency rescue at sea. However, due to limitations of time and effort, knowledge base should be further expanded and improved, and self-learning function remains to be studied.