Control systems engineering for optimizing behavioral mhealth interventions

Control systems engineering is a broad-based field that examines how system variables can be adjusted over time to improve important process outcomes. In recent years, control engineering approaches have been proposed as the basis for modeling and optimizing personalized, timevarying interventions in behav- ioral health. This chapter describes how control systems engineering principles, particularly system identification and model predictive control, can be applied to serve as dynamic modeling methods and optimal decision policies, respectively, for intensively adaptive interventions in behavioral mHealth applications. The role that behavioral theory plays in determining model structure and enabling semi- physical system identification is explained. The combined system identification-model predictive control strategy is illustrated with examples of interventions for fibromyalgia, smoking cessation, and enhancing physical activity.