On an Integrated Approach to Networked Climate Control of a Smart Home

Climate control of thermal spaces or zones is very important for complex systems like smart home. The climate control system is connected to network for the transfer of measurement data and control action packets from sensors to controller and from controller to actuators, respectively. The system can, therefore, be categorized as a cyber-physical system (CPS). Heterogeneous nature of control and cyber domains poses a great challenge in dealing with CPS development. An integrated framework of intelligent control and communication is presented in this paper for performance improvements in the climate control system. The joint framework considers relevant system objectives based on system states and actuator actions. The constraints related to errors and delays in communication along with the limited capabilities of the devices are also taken care of. The formulated problem has been solved through the real-time optimization approach following the communication protocol using two separate controller methodologies: 1) learning-based proportional-integral (PI) controller and 2) adaptive critic-based controller. The gradient descent algorithm updates the parameters of the PI controller, whereas a properly trained adaptive critic controller generates real-time control actions for achieving desired states. The real-time performance obtained for both the controllers are significantly improved even in inconsistent data communication.