Research on the Identification and Intervention System of Sense of Active Control (SOA) Based on Deep Learning

This paper proposes a deep learning-based sense of active control (SOA) identification and intervention system, which aims to identify the control perception of individuals in the process of performing tasks in real time through deep learning models, and provide customized interventions based on the identification results to enhance the individual's sense of control experience. SOA is the self-perception of an individual's control over his or her own behavior, and its level plays an important role in cognitive psychology and behavioral decision-making. This study combines physiological signals (such as electroencephalogram (EEG), heart rate variability (HRV), and galvanic skin response (EDA)) with behavioral data (such as hand movements and operation task performance) for feature extraction, and identifies SOA through DNN and CNN. Then, a personalized intervention mechanism is designed, and real-time feedback and dynamic adjustment strategies are used to optimize the sense of control experience. To verify the effectiveness of the system, this paper conducts multiple experimental simulations. The experimental results show that the SOA identification model based on deep learning can accurately identify different levels of control perception, with an accuracy rate of more than 90%. The SOA level of individuals has been significantly improved through the introduction of intervention strategies, especially under high-load task conditions. The intervention effect is more significant.