Cardiff University | Prifysgol Caerdydd ORCA
Online Research @ Cardiff 
WelshClear Cookie - decide language by browser settings

Investigation into UK public attitudes and social interactions towards electric vehicle purchasing through the integration of supervised machine learning and agent-based modelling

Xu, Wen 2022. Investigation into UK public attitudes and social interactions towards electric vehicle purchasing through the integration of supervised machine learning and agent-based modelling. PhD Thesis, Cardiff University.
Item availability restricted.

[thumbnail of PhD thesis_C1651064_Wen XU_20230710.pdf] PDF - Accepted Post-Print Version
Restricted to Repository staff only until 7 July 2024 due to copyright restrictions.

Download (5MB)
[thumbnail of Cardiff University Electronic Publication Form] PDF (Cardiff University Electronic Publication Form) - Supplemental Material
Restricted to Repository staff only

Download (248kB)


Investigating consumers’ attitudes towards electric vehicle (EV) purchasing is crucial for understanding the EV private adoption. Aggregated models may be applied to evaluate EV market penetration in a top-down approach. However, they fail to reflect individual consumers’ decision making in social environments, leading to either over-pessimistic or over-optimistic expectations. These have a negative impact on the promotion of EVs. Taking consumer heterogeneity and social influence into consideration, a new model was built, incorporating agent-based modelling (ABM) and supervised machine learning (SML) methods, to study UK consumers’ attitude formation regarding EV purchasing, using multiple social theories, in a bottom-up approach. ABM and SML were employed to simulate consumers’ social interactions and decision-making activities. Consumers’ attitude formation was determined by their socio-demographics and beliefs, based on the theory of planned behaviour; consumers’ social interactions were devised based on Rogers’s theory of diffusion of innovations and social influence theory. Then, the model was validated and optimised using parameter variation methods. The model was applied to estimate UK consumers’ attitudes and investigate potential influencing factors. This work provides a comprehensive framework to study consumer behaviour and innovation diffusion with numerical modelling, which is potentially transformative for other innovation research. The experimental results show that consumers’ attitudes towards EV purchasing are influenced by (1) the subjective norm from discussion of EVs with close friends, and (2) the provincial norm from observation of EVs in residential areas. The results indicate that consumer heterogeneity plays an important role in their decision making, underlying their attitude formation under social interactions. In addition, the results show that enhancing public environmental awareness encourages positive attitude at the early stage of EV adoption. These findings provide references for policy making to accelerate EV adoption (e.g. encouraging environmental protection), as well as new methods of EV marketing (e.g. targeting different consumer segments).

Item Type: Thesis (PhD)
Date Type: Completion
Status: Unpublished
Schools: Business (Including Economics)
Subjects: H Social Sciences > H Social Sciences (General)
Uncontrolled Keywords: consumer attitude, consumer belief, consumer heterogeneity, decision making, social interaction, electric vehicle, agent based modelling, supervised machine learning, multi-theoretical framework, multi-method modelling
Date of First Compliant Deposit: 11 July 2023
Last Modified: 11 Jul 2023 12:43

Actions (repository staff only)

Edit Item Edit Item


Downloads per month over past year

View more statistics