Research On Multi-Vehicle Path Optimization Of Carpooling Service Considering Individual Preference

With the rapid development of the urban economy and the rapid expansion of people's travel range,urban passenger transport demand is also increasing.However,due to the limitation of the total number of taxis and the influence of the organization of operations,the imbalance between transportation supply and demand has led to the emergence of problems such as "difficulty in riding",which has seriously hindered the sustainable development of urban transportation.Vehicle sharing is a way of effectively relieving traffic pressure.It has huge market demand,can improve the carrying efficiency of vehicles,integrate transportation resources,and save travel costs.In order to improve the efficiency of carpooling service,and meet the personalized service needs of more passengers,this paper comprehensively analyzes the economy,timeliness,sociality and safety in the passenger ride experience,taking into account the individual preferences of the passengers,the passenger individual preference demand indicator matrix is determined,the travel utility under the non-multiplicative and multiplication modes is calculated based on the passenger utility,and the discrete utility model is used to evaluate the passenger satisfaction value.At the same time,taking into account passengers' willingness to ride,baggage carrying,time window,car capacity and other factors,balancing the interests of both the drivers and the passengers,with the highest passenger satisfaction,the highest driver's income,the lowest vehicle carrying cost,and the lowest vehicle idle cost.A rise-sharing matching and route optimization model for allowing single-pass and multi-share coexistence is established for different types of passengers.Aiming at the characteristics of ride-sharing matching and path optimization model,this paper designs an adaptive hybrid genetic optimization algorithm to solve the model.This optimization algorithm can adaptively adjust the crossover and mutation probability to influence individual changes according to the state change of the population as a whole.In addition,this paper builds a multi-vehicle matching and path optimization simulation system that allows multi-mode coexistence based on Matlab development environment.The test case simulation results show that the multi-vehicle match based on individual preference is compared to a fixed match the system driver revenue can be increased by 2.6%,the average passenger satisfaction is increased by 11.8%,the comprehensive income increased by 7.4%,and the matching success rate increased by 15.0%.At the same time,increasing the proportion of willingness,adjusting the size of the fleet,and increasing the maximum tolerated distance of passengers can improve the carpooling effect.The research in this paper enriches the theoretical research on the problem of the multiplication path optimization.At the same time,it can provide reference and suggestions for traffic managers and taxi operators in practical applications,and promote the healthy development of urban transportation.