Intelligent Scheduling Systems For Public Vehicles In Smart Cities

Electric vehicles begin in 1830 s,which receive attention for low energy consumption and less pollution.Driverless car technology begin in 1920 s,which is a main field of artificial intelligence,and can sense the road environment by vehicular sensors and arrive at destinations through path planning strategies.Ridesharing begin in the early20 th century in North America,and after the oil crises in 1970 s becomes popular for less dependency on energy.From the above tendencies of technologies,we can predict that,a new transportation,named as public vehicle(PV)system,will become true in smart cities based on the driverless car technology,electric vehicle,ridesharing and mobile Internet.Its execution process is as follows: Passengers send trip requests to the cloud including the earliest start time,origins,destinations,etc.The cloud calculate new paths after getting the requests,and then assign corresponding PVs(there may exist passengers inside)to serve passengers,and at the same time the paths of PVs will be updated.In the development of the PV system,this paper focuses on the smart scheduling systems based on the vehicle hardware.Strictly speaking,the following problems are solved:1.A path planning approach based on dynamic ridesharing.The paths of PVs will change if new requests enter.The path planning problem includes two parts.First,determine the matching between PVs and requests.Second,the paths should be calculated again to serve new requests.As the PV system provides dynamic ridesharing service,which has different characteristics from traditional ridesharing.This improves the complexity of the algorithm design.At the same time,these requests should be handled by the cloud in real-time.The experiments based on Shanghai show that,the achieve the same performance(e.g.,the total travel time)the number of PVs is much smaller than conventional vehicle system and conventional ridesharing system.The traffic efficiency is improved.2.Path planning improvement approaches based on search and transfer.The traffic big data based path planning gives new challenges for the PV system.It needs more efficient search algorithms to satisfy the real time demand.This paper proposes a big data based path planning approach to considering of the above requirements.This approach not only can reduce the amount of computation,but also ensure the QoS of passengers.To improve the feasibility of PVs,enhance the coordination of PVs,we propose a transfer approach,so that passengers can transfer between different PVs.3.Coordination approach for transportation and charging based on game.The main functions of PVs are transportation(providing trip services)to relieve traffic pressure.At the same time,PVs are one type of electric vehicles,which need to obtain energy from smart grids.To balance the relationship between transportation and charging,we build a game model where PV groups adjust their transportation and charging strategies to maximize their utilities.This paper proposes an approach to balance the transportation and charging demands of PVs.The aim is to reduce the traffic pressure,obtain low-cost energies,and finally ensure the long term operation of the PV system.Combining the above three studies,we construct the whole research for the PV system.