Abstract:

Agent-based modeling (ABM) offers a nuanced framework for representing the bounded rationality of individual decision-makers. But in order to be useful ABM applications in human-technical systems demand rigor and robustness regarding empirical grounding and theoretical orientation. In this paper we present the architecture of a theoretically-based and empirically-driven agent-based model fot technology adaptation, with an application to residential solar photovoltaics (PV).  Focusing on aspects of model design, setup, initialization, and validation, we present a comprehensive approach for integrating empirical data into the initialization of agent states, while framing agent decision rules based on theory.  This integrated ABM framework is applied to real-world data between 2004-2013 for a residential solar program at the city scale, with household-level resolution for demographic, attitudinal, social network, and environmental variables.