Application Of Type-2 Fuzzy Sets In Driving State Evaluation And Behavioral Decision Of Unmanned Vehicle

As part of the intelligent transportation system,the smart car will greatly reduce traffic safety accidents and alleviate traffic congestion.The ultimate form of a smart car is unmanned,and behavioral decision-making is the core of unmanned driving.The decision-making rules will directly affect the safety of tnmanned cars and the smoothness of road traffic.So,it is especially important to choose the right method and develop a set of effective decision-making rules for unmanned cars.Firstly,by summarizing the structure of the automatic driving control system and the decision-making layer of the unmanned vehicles,the whole working principle of the unmanned driving is summarized and the fuzzy control method is applied to the decision-making rules of the unmanned car and further a set of decision-making rules is designed.For the input of the control system of unmanned cars,the paper first summarizes the factors that may affect the safety of the driving vehicle;in order to evaluate the driving safety of the unmanned vehicle by using fuzzy comprehensive evaluation method,it then determines the "driving speed","relative speed" and "interval distance" as the evaluation indicators which are expressed in the form of interval type-2 fuzzy sets;finally,according to the driver's experience and actual situation,based on the type-2 fuzzy comprehensive evaluation method,this paper gives a method to evaluate the driving safety of unmanned cars.Finally,according to the rule base of the unmanned vehicle control system,based on the basic principles of driving and safe distance,the paper uses fuzzy reasoning theory to develop a fuzzy rule base of behavior decision making that can be applied in various traffic scenarios.The evaluation result of its individual indicators can be used as the input of fuzzy decision rules and the validity of the rules is verified by the fuzzy logic toolbox of Matlab.All of tihese together have provided a new idea for the research of the decision-making technology of the unmanned cars.