Abstract
Intelligent Transportation Systems (ITSs) can be applied to inform and incentivize travellers to help them make cognizant choices concerning their trip routes and transport modality use for their daily travel whilst achieving more sustainable societal and transport authority goals. However, in practice, it is challenging for an ITS to enable incentive generation that is context-driven and personalized, whilst supporting multi-dimensional travel goals. This is because an ITS has to address the situation where different travellers have different travel preferences and constraints for route and modality, in the face of dynamically-varying traffic conditions. Furthermore, personalized incentive generation also needs to dynamically achieve different travel goals from multiple travellers, in the face of their conducts being a mix of both competitive and cooperative behaviours. To address this challenge, a Rule-based Incentive Framework (RIF) is proposed in this paper that utilizes both decision tree and evolutionary game theory to process travel information and intelligently generate personalized incentives for travellers. The travel information processed includes travellers’ mobile patterns, travellers’ modality preferences and route traffic volume information. A series of MATLAB simulations of RIF was undertaken to validate RIF to show that it is potentially an effective way to incentivize travellers to change travel routes and modalities as an essential smart city service.
Highlights
If we are to achieve more beneficial and cohesive personal, business and societal goals, in the face of increasing global challenges such as an increasing sedentary lifestyle for citizens, the promotion of physical environment sustainability, a reduction in poverty, better physical and social living conditions, etc., we need to change our behaviour and consider how to do this more effectively
To solve the personal incentive problem, we can weight each criteria to represent the importance of a criterion to a traveller, which is according to the sensitivity of the traveller to the criterion using Analytic Hierarchy Process (AHP)
The usability of the incentive decision tree is highly related to the design of Iterative Dichotomiser 3 (ID3) and the historical sample data of all the objective travellers used in the training
Summary
If we are to achieve more beneficial and cohesive personal, business and societal goals, in the face of increasing global challenges such as an increasing sedentary lifestyle for citizens, the promotion of physical environment sustainability, a reduction in poverty, better physical and social living conditions, etc., we need to change our behaviour and consider how to do this more effectively. We often tend to have more than one personal transport use goal [10], so we need to consider how we need to more effectively design multi-dimensional incentives, e.g., ones that use multiple modes and that are context driven, as well as personalized ones with respect to travellers’ preferences and location and the real-time travel situation. The use of machine learning and game theory has been proposed as a promising candidate to solve the computing intelligence and multi-dimensional goal problem [15,16] Following this trend, we designed a Rule-based Incentive Framework (RIF) to intelligently generate context-driven incentives for travellers, while considering the multi-dimensional goals. Propose a rule-based incentive framework to solve the personalized context-driven incentive problem, whilst achieving personalized travel preference, travel efficiency and eco-friendly global transportation as the multi-dimensional goals.
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
