Online Identification and Temperature Tracking Control for Furnace System With a Single Slab and a Single Heater Over the Wirelessly Connected IoT Controller

The swift evolution of Internet of Things (IoT) technologies has facilitated the rapid deployment of heterogeneous devices in industrial systems. In this work, we focus on the identification and temperature tracking control for a furnace system with a single slab and a single heater over a wirelessly connected IoT controller. Specifically, we characterize the IoT-based furnace temperature control system using an integrated-control-and-communication framework that involves thermal modeling of convection and conduction in the furnace plant, as well as wireless modeling of the communication network between the furnace plant and the remote estimator-controller center. Based on the framework, we first propose a novel stochastic-approximation-based online algorithm to learn the optimal temperature tracking control solution for the furnace system with knowledge of the furnace dynamics. After that, we extend the temperature tracking control approach to deal with the furnace system with unknown furnace dynamics and propose a novel normalized-stochastic-gradient-descent (NSGD)-based algorithm to simultaneously identify and control the furnace system at the remote controller in an online manner. Using the Lyapunov stability analysis and ordinary differential equation (ODE) method, we theoretically demonstrate the asymptotic convergence of the proposed learning algorithms. Numerical analysis is conducted for our proposed temperature tracking control scheme and several state-of-the-art schemes. Our results demonstrate that our proposed scheme outperforms the baseline schemes in terms of temperature tracking accuracy and fuel efficiency, in the presence of the wireless interface.