Building Control


Approximately, the buildings consume about 20-40% of the global energy, around half of which is used for heating, ventilation, and air conditioning system (HVAC). Therefore, it is necessary to consider advanced HVAC control strategies in order to reduce energy consumption and satisfy individuals’ thermal comfort level demand at the same time. During the controller design process, both the outside environment disturbance and inside human activities should be considered. Although weather forecast and human activity schedule estimation could provide basic information of the disturbance, uncertainties still cannot be avoided. In addition, there are many constraints imposed by practical engineering system that need to be handled, including saturation constraint, stability constraint and so on.

Research Topics

Iterative design of feedforward and feedback controller

Note that for most commercial or residential buildings, there always exists some certain repetitive patterns for disturbance, including outside environment and inside human activities. With the repetitive nature of the disturbance, iterative learning control (ILC) is a suitable solution to consider. The central idea of ILC is to improve the performance of the current iteration based on learned information from previous iterations. Therefore, repetitive disturbance can be removed. In order to reduce the influence of non-repetitive components at the same time, an iterative tunable feedback controller is introduced. The control problem is formulated as an optimization problem. Through adjusting feedforward ILC signal and feedback controller iteratively, the control performance can be obviously improved.

ILC with IFT design
HVAC System Overview


  1. C. Peng, W. Zhang and M. Tomizuka, “Iterative design of feedback and feedforward controller with input saturation constraint for building temperature control,” 2016 American Control Conference (ACC), Boston, MA, 2016, pp. 1241-1246.
  2. C. Peng, L. Sun, W. Zhang and M. TOmizuka, “Optimization-based constrained iterative learning control with application to building temperature control,” 2016 Advanced Intelligent Mechatronics (AIM), Banff, Canada, 2016.


Cheng Peng Graduate Student Email Link  
Shuyang Li Graduate Student Email Link