Research Of Validation Method Of Emergency Deduction Rules Based On Knowledge Element

With the rapid development of social economy,the living environment of human is being more and more bad,followed by the frequent of all kinds of emergencies,such as tsunamis,earthquakes,landslides,fire,etc,which bring human society enormous pressure of survival and property losses.The relevant government departments and scholars at home and abroad have given sufficient attention to them,which makes emergency decision-making become an important research direction.At present,“scenario-response” model have been widely recognized by emergency field scholars and experts gradually,then rule-based inference as the main way of emergency situation evolution have also been widely used.With emergencies occurred repeatedly,emergency cases and inference rules' database have been repaired constantly,which has brought huge challenge to the effective use of the emergency cases and the consistency of the inference rules' database.At the same time,the emergency inference efficiency is not high enough so that also puts forward the requirements for the applicability of the rules.Therefore,how to validate the consistency of emergency rules' database and promote the correctness and applicability of inference rules is one of the key problems of the guarantee of the inference based on rule and emergency decision.In order to improve the consistency of emergency inference rules' database,the utilization rate of emergency cases and the applicability and accuracy of emergency inference rules,this paper proposes a method of validation the structure errors of inference rules by using Petri nets and its relationship matrix and validates inference rules with similar cases.Firstly,it takes the general knowledge model as a foundation of scenarios,common inference rules and cases,then expresses them respectively in turn;After that,it turns the expression of inference rules to Petri nets and points out the existing method of validation of the structure errors of inference rules,then optimizes its definition.Also it proposes a method of validation the structure errors of inference rules by using relationship matrix of Petri nets.And the effectively of the method is proved by the examples which can ensure the consistency of inference rule database for the subsequent applicability of inference rules.Finally,this paper uses the current scenario to match emergency inference rules,then extracts the matched inference rules,After that those extracted inference rules combine with the current scenario to match emergency cases,and it applies those matched cases to verifying extracted inference rules by deduction,the valid inference rules are used in scenario deduction to provide decision support for scenario deduction of emergency.Applying the debris flows happened in Qipan gully of Wenchuan and other places to examples validated the effectiveness of the method.In order to keep the consistency of inference rule database,the structure errors of emergency inference rules' database has been validated in this paper,thus it provides a favorable rule basis of the validation rules applicability.And it uses the similar cases to verify the inference rules in the process of the validation rules applicability,which promotes the utilization of cases and the applicability and accuracy of inference rules,as a result it can carry out emergency decision more smoothly.