Modeling and optimization for integrated people-and-goods transportation systems

Abstract

Urban (goods) delivery and passenger transport are essential components of urban infrastructure. In recent years, an integrated people-and-goods transportation model that utilizes passenger transport vehicles for goods deliveries has been touted as one of the most promising alternatives to optimize both goods delivery and passenger transport. Nonetheless, several implementation challenges may hinder the viability of this service model. This thesis aims to address the core challenge of pricing optimization, along with a series of strategic- and operational-level problems for three specific forms of integrated people-and-goods transportation service. Study I in Chapter 3 introduces an ordinary traveler-based crowd-shipping service, addressing a compensation and service routing (C&R) problem for a hybrid delivery system combining a dedicated delivery service and a crowd-shipping service by ordinary travelers. Two mixed integer programming models are formulated for the C&R problem under uniform and differentiated compensation modes. A customized hybrid algorithm, which employs a variable neighborhood search with a nested tabu search and an iterated local search, is developed to solve the problems. Study II in Chapter 4 extends the investigation in Chapter 3 by examining a collaborative alliance, allocation of delivery orders, compensation, and route joint optimization (CACR) problem for a collaborative hybrid delivery system involving multiple retailers. A bi-objective optimization model is formulated to minimize total operational costs and carbon emissions. A decomposition-based iterative optimization method is developed to find Pareto-optimal solutions by solving a series of decomposed sub-problems with the updated collaboration strategy and compensation rate. Each sub-problem is solved by using a cluster-first route-second approach with a customized spatiotemporal clustering technique and a non-dominated sorting genetic algorithm-II enhanced with a Clarke and Wright saving method. Study III in Chapter 5 investigates a public transit-based co-modal transportation service price (CSP) problem considering the collaborative interaction between the logistics service provider (LSP) and public transit operator (PTO). A bilevel path-based programming model is formulated based on the interactive dynamics between LSP and PTO, where a lower-level bus trip scheduling problem is proposed for optimizing the PTO's decision while an upper-level vehicle routing problem with pricing is designed for optimizing the LSP's decision. A tailored iterated three-stage hybrid method, combining two granular tabu search algorithms and an artificial bee colony algorithm, is developed to solve the problem. Study IV in Chapter 6 examines an outsourcing service price (OSP) problem for co-modal delivery service based on on-demand mobility services. A lower-level co-modality delivery service problem with ridesharing is formulated to determine the OMP's optimal decision under the outsourcing service price offered by PSP, while an upper-level multi-depot pickup and delivery problem with pricing is formulated to determine the PSP's optimal decision, including the service price. A customized iterative hybrid algorithm, integrating two granular tabu search algorithms and a genetic algorithm, is developed to solve the problem. These mathematical models and solution methodologies formulated in the four studies are tested on adapted benchmark instances, randomly generated instances, and real-life cases, and offer managerial insights for the urban delivery service providers.