Modelling reservation-based shared autonomous vehicle services: A dynamic user equilibrium approach

Shared Autonomous Vehicles (SAVs) are expected to be used for regular and pre-planned trips. Such trips suitable reservation-based services, wherein the customer needs book a trip in advance. Systems enabling reservation of can allow better planning routes schedules, if optimally designed, enable higher efficiency. The primary objective this research is model effects such system, by formulating solving combined Dynamic User Equilibrium autonomous vehicle Chain Formation (DUESCF) problem. problem formulated as bilevel based on game theory, involving road users SAV service operator. Given situation where conventional private shared vehicles co-exist, select paths departure times maximize perceived utility (commonly treated minimizing disutility) forming DUE (fixed point problem), operator tries performance appropriate chains (combinatorial problem). final traffic assignment that includes chain formation, both obtain optimal solutions reaching Nash equilibrium, no player off unilaterally changing their decisions. A solution approach, Iterative Optimization Assignment (IOA) method, proposed with path flow changes convergence criteria. Furthermore, approach tested its robustness, scenario analysis carried out evaluate impacts services. results show ridesharing system compared carsharing mixed consisting both, terms total travel time, congestion levels, kilometres travelled requirements. • This study integrates fields transport modelling, operations theory. An extended developed SAVs. Interaction between an modelled equilibrium. Experiments conducted ensure stability robustness algorithm. Scenario shows result positive impacts.