Study On The Location Allocation Problem Of Vehicle Inspection Station In Uncertain Environment

Vehicle inspection is necessary for the technical status or working ability of the vehicle.Timely vehicle detection can ensure that the vehicle has good safety,reliability and environmental protection.In recent years,with the rapid development of economy and the rapid increase of vehicles,the demand for vehicle inspection is increasing progressively.Many cities have to build new Inspection stations to meet the new inspection demand.Location allocation is the first step in the construction of inspection stations and he first step of location allocation is to analyze customer needs.At present,the deterministic method is mainly used to predict the customer needs.However,the custom needs which can not be ignored will change with the change of Vehicle ownership ? region ? and inspection station number in the actual location process.Therefore,it is necessary to study the location allocation of vehicle inspection station.Uncertain environment includes random environment,fuzzy environment,fuzzy random environment and random fuzzy environment.Random environment means that some areas have the historical data of inspection customers,which can be used to fit the probability distribution of requirement.According to this,the model that can locate the vehicle inspection station in random environment is studied in my dissertation.To solve this problem,considering the custom's demands as random variable,the dissertation constructs a single-objective location model and a multi-objective location model.The single-objective location model aims at testing the lowest total cost of customers,and constrains the income of investors and the geographical range of the location of inspection stations.The multi-objective location model that aims at testing the lowest total cost and the shortest travel time of customs can be treated the expected value?confidence level and geographic range as constraints.A model solving method combining random simulation and "teaching and learning" optimization algorithm is designed.The random simulation method and teaching and learning algorithm are used to solve the function of random variables and the position of the detection station,and the model and algorithm are tested by case.Fuzzy environment means that no data some area or the data is not enough,it is not possible to estimate the customer requirement by probability distribution.In this case,the requirement can be determined by consulting experts.Based on the fuzziness of requirement,both single-objective and multi-objective location model under the fuzzy environment are proposed.To solve the problem,The single-objective location model aims at the lowest total cost of customs,and constrains the income of investors and the geographical range of the location of inspection stations.The multi-objective location model aims at the lowest total cost and the shortest travel time of customs,which takes the expected value?confidence level and geographic range as constraints.A model solving method combining fuzzy simulation and Artificial Bee Colony algorithm is designed.The fuzzy simulation method and Artificial Bee Colony algorithm are used to solve the function of fuzzy variables and the position of the detection station,and the model and algorithm are tested by case.Fuzzy random environment or random fuzzy environment refers to random and fuzzy that are co-existent in one case.For example,the customer requirement is a random variable,different experts may give different characteristic parameter values for the random distribution.In this case,it is not accurate to describe it as a random or fuzzy variable.It can be described as a random fuzzy variable.In the same way,in the fuzzy situation,some of the parameters may be conformed to be a certain probability distribution.It can be described as a fuzzy random variable in this situation.Therefore,considering the shortage of random or fuzzy,the dissertation proposes the location model under the co-existence of random and fuzzy innovately.Fuzzy random theory and random fuzzy theory are introduced into the problem.In the case of fuzzy random environment,the number of customers is considered as the fuzzy random variable.Taking the lowest total cost of customers as the goal and taking the expected value?confidence level and geographic range as constraints to construct the model.In the case of random fuzzy environment,the number of customers is taken as the random fuzzy variable.Taking the lowest total cost of customers as the goal and taking the expected value?confidence level and geographic range as constraints to construct the model.A model solving method combining uncertain simulation and intelligent algorithm is designed.The uncertain simulation and intelligent algorithm are used to solve the function of uncertain variables and the position of the detection station,and the model and algorithm are tested by case.Using uncertainty theory to analyze customer demand is more consistent with the growth law of customer demand.The research that enriches the location model and solving method of vehicle inspection station under uncertainties and provides theoretical guidance and decision-making is supported for the planning practice of vehicle inspection station.The major innovations include:(1)A stochastic multi-objective location model considering regional constraints is constructed,and "teaching and learning" optimization algorithm is designed to solve the model.(2)Under less data of customer demand,a fuzzy programming model for vehicle detection station location is established,and an artificial bee colony algorithm is designed to solve the model.(3)The location model of vehicle detection station under the coexistence of fuzziness and randomness is proposed,and the particle swarm optimization(PSO)algorithm is designed to solve the random fuzzy allocation model.