Research And Application Of The Evaluation Expert System Based On Ontology For Road-driving Skill Examination

As an application-oriented research field of artificial intelligence (AI), expertsystem (ES) has been widely applied to many fields of Chinese national economy. Inrecent years, with the rapid development of the technologies of AI, especiallyOntology, the concept and implementation techniques of ES are being changed.Although both ES based on rules and traditional programs based on algorithmcouldbe used to solve certain problems, ES has incomparable superiority than traditionalprogram.With the development of Chinese national economy, more and more people takepart in the motor vehicle driving test in order to get a driving license. In recent years,the Road Driving Test System is applied widely in the test of driver qualification toincarnate the scientificity and fairness by many DMVs (Department Of MotorVehicles) in our country. The existing Road Driving Test Systems, or RTS3for short,have been almost developed with traditional programming methods based onalgorithm. There are two shortcomings for the RTS3systems. Firstly the RTS3systems are poor to maintain. Secondly, because there are no unified, sharedevaluation rules of auto-judging items for the RTS3systems, the results of thedifferent RTS3systems to evaluate the same auto-judging items may bevarious.Building the RTS3system based on ES technology can greatly improve themaintainability of the system. Applying Ontology into the RTS3systems can make iteasy to share and exchange the evaluation rules among the different systems. In theend, a complete, unified library of evaluation rules for the road-driving skillexamination will be formed.So,The research objects of this thesis arethe expert system model based onOntology and implementation of the RTS3system based on Ontology.The main work and reserach of this thesis are as follows.First, for the maintenance of the RTS3system and sharing the evaluation rulesamong the systems, AI, especially Ontology, is introduced to the RTS3system. The evaluation module of the system is designed as an expert system. So the evaluationexpert system based on Ontology for road-driving skill examination is constructed.Second, ESBO, short for an expert system model based on Ontology, reasoningrule and its sequence,is proposed. The implementation of the model is discusseddeeply, such as the method of ontology representation, ontology reasoning, andrulesrepresentation. The inference process based on OWL and Jess are alsointroduced. The ES model based on the OWL, SWRL and Jess is designed byreferencing the ESBO. The reasoning model based on the JessML and OWL isproposed. Three ways are proposed to improve the efficiency of the inferencemachine. The first one is to extend OWL’s modeling primitive by adding the<owl:Sequence> element. The second is to follow the rules: FFAF, VFAL andMMSAF aswriting SWRL-based rules. The third is to search the inferred facts byusing a new efficient algorithm (RRS).Third, the concept model of the domain ontology of the evaluation expertsystem for road-driving skill examination, or RTOM for short, is given after thedriver test system and related concepts are analyzed. The ontology library of theevaluation expert system for road-driving skill examination is built in OWL language.Based on the ontology library, the auto-judging rules library of the evaluation expertsystem is built in SWRL language. They are generated in Protégé by using OWL andSWRL language.Fourth, the model of the evaluation expert system based on the OWL, SWRLand Jess, is designed by referencing the ESBO and the ontology library andauto-evaluation rules library. An implementation of the expert system based on Jess,C#and Java is introduced in detail.Based on the principle of Ontology, the expert system model based on Ontologyis studied. The ontology library of the evaluation expert system for road-driving skillexamination is built. Also the implementation techniques of the evaluation expertsystem are discussed deeply. All the results of these studies provide some universaltheoretical models, development schemes and implementation techniques forresearching and implementation of the ES based on Ontology. To promote thestandardization and normalization of the driving road test system, the ontology libraryand rules library of the evaluation expert system for road-driving skill examination are given. The ontology library and rules library are easy to share and exchangeamong the RTS3systems.The research work in this thesis was supported by the second sub-topic of thefifth topic of the National Science and Technology Action Plan for Road TrafficSafety (2009BAG13A05) made by the Ministry of Science and Technology P.R.C,the Ministry of Public Security Ministry of Public Security P.R.C and the Ministry ofCommunications P.R.C, namely researching of the evaluation technology of theDriver Road Examination System and developing of the system.