Research On Scenario Deduction And Emergency Knowledge Matching Of Fire Accident In Chemical Industry Zone

The rapid development of chemical industry at home and abroad triggers a construction and development boom of chemical industrial zone in nationwide,however chemical industrial zone is easy to cause all kinds of accidents because of its production characteristic.According to statistics,there are more than 50 chemical industrial zone accidents occurred only in 2015,in which there are 35 fire accidents,accounting for about 70%.Once fire accident happens,it is important to determine its development situation accurately and make effective emergency decision-making for rescue facing the ever-changing accident.So the research uses the idea of "scene-response",studying on scenario deduction and emergency knowledge matching method on the basis of the studying on situation inference in order to build a dynamic decision support method for fire accidents in chemical industrial zone,providing assistant decision-making information for decision-makers and minimizing the damage of fire accidents.The study analyzes the scenario elements of chemical industrial zone fire accident and describes the current situation based on frame representation method at first,representing the whole accident with the form of nested framework,combing the logical relationship between different frameworks.Then the next moment scene is deduced dynamically on the basis of the current situation with the view of evolution and derivative law in accidents.Scenario properties are classified according to different features.The development trend of group properties is deduced using objective cellular automata method and subjective & objective trend prediction method for the randomness Based on the current and forecast scenario,the study uses semantic similarity and accident evolution degree as measurement index and studies on knowledge matching with case-based reasoning method.According to the characteristics of the scenario property,different distance algorithms are used for different groups and corresponding auxiliary decision-making information is given finally.