Strategic Trip Planning: Striking a Balance Between Competition and Cooperation

Abstract

In intelligent transportation systems, cooperative mobility planning is considered to be one of the challenging problems. Mobility planning as it stands today is an in- dividual decision-making effort that takes place in an environment governed by the collective actions of various competing travellers. Despite the extensive research on mobility planning, a situation in which multiple behavioural-driven travellers partic- ipate in a cooperative endeavour to help each other optimize their objectives has not been investigated. Furthermore, due to the inherent multi-participant nature of the mobility problem, the existing solutions fail to produce ground truth optimal mobil- ity plans in the practical sense - despite their claimed and well proven theoretical optimality.

This thesis proposes a multi-module team mobility planning framework to address the team trip planning problem with a particular emphasis on modelling the inter- action between behaviour-driven rational travellers. The framework accommodates the travellers’ individual behaviours, preferences, and goals in cooperative and com- petitive scenarios. The individual behaviours of the travellers and their interaction processes are viewed as a team trip planning game. For this game, a theoretical anal- ysis is presented, which includes an analysis of the existence and the balancedness of the final solution.

The proposed framework is composed of three principal modules: cooperative trip planning, team formation, and traveller-centric trip planning (TCTP). The cooper- ative trip planning module deploys a bargaining model to manage conflicts between the travellers that could occur in their endeavour to discover a general, satisfactory solution. The number of players and their interaction process is controlled by the team formation module. The TCTP module adopts an alternative perspective to the individualized trip-planning problem in the sense that it is being behavioural driven problem. This allows for multitudes of traveler centric objectives and constraints, as well as aspects of the environment as they pertain to the traveller’s preferences, to be incorporated in the process. Within the scope of the team mobility planning frame- work, the TCTP is utilized to supply the travellers with personalized strategies that are incorporated in the cooperative game. The concentration problem is used in this thesis to demonstrate the effectiveness of the TCTP module as a behavioural-driven trip planner.

Finally, to validate the theoretical set-up of the team trip planning game, we introduce the territory sharing problem for social taxis. We use the team mobility framework as a basis to solve the problem. Furthermore, we present an argument for the convergence and the efficiency of a coarse correlated equilibrium. In addition to the validation of a variety of theoretical concepts, the territory sharing problem is used to demonstrate the applicability of the proposed framework in dealing with cooperative mobility planning problems.