Analysis Of Mechanism And Control Strategy Of Route Choice And Departure Time Choice In Autonomous Mobility And Ridesharing

The past decade brought two major innovations to transportation: connected and autonomous vehicles(CAV)and transportation network companies(TNC)such as Uber and Lyft.The transformative impacts of these innovations are expected to continue to reshape the industry.Both the academia and the industry expect these two innovations to improve the safety of transportation and effectively reduce traffic congestion.However,since these two innovations can significantly improve the convenience of travel,it may induce more travel demand and consequently further worse traffic congestion.According to a recent study published in the journal of Science Advance,the ride-hailing companies like Uber and Lyft did make San Francisco's traffic worse.Hence,we should take new technology and new scenarios together to ensure positive impact of the proposed implementations that may arise in the future.In order to develop effective traffic control measures and travel demand management measures,we must have a deep understanding of travel rules and their characteristics.The paper is organized as follows.Chapter 2 and Chapter 3 focus on the analysis of the travel rules of urban commuters,and then Chapter 4 and Chapter 5 expound that how to optimize the performance of the transportation system via automated vehicles and ridesharing.The main contents and contributions are summarized as follows.(1)A day-to-day dynamic model with user learning is built via fractional calculus approach.Compare to existing day-to-day dynamics that are all characterized via ordinary differential equations,fractional differential equations are used to model travelers' day-to-day route swapping behavior.Moreover,what is different from other learning day-to-day dynamics is that the learning behavior obeys a power law decay instead of exponential decay law.By analogy with a fractional oscillatory system,the generalized kinetic energy,the potential energy and the total energy of the traffic system are well defined.Inspired by the Minimum Total Potential Energy Principle,we prove that the total energy can be a Lyapunov function and then the system asymptotically converges to user equilibrium.(2)A stochastic day-to-day model and a general Lyapunov function are established to capture the perceptual errors in travelers' day-to-day route choice behavior.Inspired by the analogy between stochastic user equilibrium and a mixture in thermodynamics,we show that Fisk's formulation can be interpreted as the synonymous “Gibbs free energy” of a transportation network and then define the “chemical potential” of route.The stochastic user equilibrium is obtained when the “chemical potential” of each route is equal and the Gibbs free energy reaches minimum.We built a stochastic day-to-day model based on routes' chemical potential.Inspired by rational behavior adjustment process,we first define the stochastic rational behavior adjustment process(SRBAP).As the counterpart of the Beckmann transformation,which has been widely used as a candidate Lyapunov function to prove the stability of continuous day-to-day traffic evolution models that converge to deterministic user equilibrium,Fisk's formulation is utilized in our study as a general Lyapunov function for the day-to-day models that converge to stochastic user equilibrium,so far as the flow evolution satisfies SRBAP.The Logit dynamic,the Logit-based Smith dynamic,and the Logit-based Brown-von Neumann-Nash(BNN)dynamic are given as three examples under this framework.(3)A mixed day-to-day traffic model is built to manage partially automated network traffic flow.In this section,we analyze how a central agent may bring a mixed traffic system including both human-driven and autonomous vehicles to an equilibrium that both maximizes the efficiency and is stable under the control.First,the existence and uniqueness of the mixed traffic equilibrium is examined based on variational inequality equation.The evolution of the human drivers' route choices,as well as the agent's control measures,is described using a joint day-to-day(DTD)dynamical model based on probability route choice.Within this setting,we show that(1)the fixed point of the proposed dynamical system coincides with the unique mixed equilibrium,and(2)the system is asymptotically stable in continuous time,namely it always converges to the mixed equilibrium from a given initial state.We then examine how alternative control policies may affect the transition trajectory leading to the mixed equilibrium.Two alternative control schemes are proposed and analyzed.The first,referred to as the stabilityfirst control,aims to stabilize a given disequilibrium as soon as possible.The second seeks to minimize the total system cost accumulated over the transition period,hence called the efficiency-first control.We propose a continuous time optimal control formulation for both schemes and discuss how the formulation can be discretized and solved to local optimality using existing algorithms.Numerical experiments conducted on two illustrative examples highlight the differences among the three control schemes and how the share of autonomous vehicles affects the tradeoff between the efficiency and stability of the mixed traffic system.(4)A novel quantity-based travel demand management system aiming to promote ridesharing is built.The system sells the permit to access a road facility by auction in which the congestion can be eliminated.Meanwhile this system encourages commuters to share the permits with each other that further enhances bottleneck capacity.At the core of this auctionbased permit allocation and sharing system(A-PASS)include four aspects:(1)the passing time slot;(2)the ride-sharing role;(3)the matching partner;and(4)the payment.The first three questions are a trilateral matching problem(TMP)that can be formalized as a 0-1 integer programming and we prove it can be reduced to an equivalent linear program which can be solved in polynomial time.The difficulty for pricing the bottleneck permit and ridesharing is how to establish a reasonable price mechanism to satisfy the following principles: budget balance(BB),individual rational(IR),allocative efficiency(AE)and incentive compatibility(IC).A pricing policy based on the classical Vickrey-Clark-Gloves(VCG)mechanism is proposed to determine the payment for each commuter.The VCG mechanism satisfies IR,IC,AE while can't ensure BB.We also show the operator can eliminate any deficit that may arise from the application of the VCG mechanism,by controlling the number of shared rides.Results of numerical experiment suggest A-PASS strongly promote rider-share.As ride-sharing increases,all stakeholders are better off: the operator receives greater profits,the commuters enjoy higher utility,and the society benefits from more efficient utilization of infrastructure.This research will contribute the understanding of complex travel behaviors and provide a theoretical basis for refined transportation planning management,particularly on routing and departure time in autonomous mobility and ridesharing.