Abstract
Abstract
Intelligent home terminals serve as the key link between user personalized needs and the coordinated operation of home devices. Their performance directly affects the living experience, device energy efficiency, and response speed to user demands. Current control methods for intelligent home terminals face challenges such as insufficient collaborative processing of heterogeneous data sources and weak adaptability to dynamic scenarios, which limits precise control decisions in complex home environments. This study proposes an intelligent home terminal control method based on a Multi-sensor Data Fusion algorithm. It enhances the efficiency of integrating multiple data sources through the high-dimensional feature classification capability of Support Vector Machines and optimizes uncertain decision outputs using the probabilistic reasoning ability of Bayesian Networks, achieving high-precision dynamic control of intelligent home terminals. Test results show that the model achieves 98.15 % accuracy in environmental feature recognition in the training set and a control command response time of 38.51 ms. In the indicator tests, the accuracy of unstructured feature extraction reaches 98.72 %. These results demonstrate that the proposed model effectively addresses the insufficient multi-modal data fusion and weak adaptability in current intelligent home terminal control, providing a new pathway for the optimised design of intelligent home terminals.