Distributed and cooperative optimization-based iterative learning control for large-scale building temperature regulation

Cheng Peng, Liting Sun, Masayoshi Tomizuka

Abstract: In this paper, a distributed and cooperative optimization-based iterative learning control (ILC) algorithm is proposed for large-scale building temperature control problems. With the algorithm, large-scale building temperature control problems are solvable with reasonable computational load and guaranteed control performance under nearly repetitive disturbances. The large-scale centralized system is separated into several distributed and cooperative small-scale subsystems that communicate among each other. For each subsystem, a convex optimization problem is solved. The cooperative learning policy allows all subsystems to contribute together to improve the overall performance. The convergence property of the algorithm is proved and simulation results are provided to demonstrate its effectiveness.