Model predictive control with adaptive machine-learning-based model for building energy efficiency and comfort optimization

Shiyu Yang,Man Pun Wan,Wanyu Chen,Bing Feng Ng,Swapnil Dubey
DOI: https://doi.org/10.1016/j.apenergy.2020.115147
IF: 11.2
2020-08-01
Applied Energy
Abstract:A model predictive control system with adaptive machine-learning-based building models for building automation and control applications is proposed. The system features an adaptive machine-learning-based building dynamics modelling scheme that updates the building model regularly using online building operation data through a dynamic artificial neural network with a nonlinear autoregressive exogenous structure. The system also employs a multi-objective function that could optimize both energy efficiency and indoor thermal comfort, two often contradicting demands. The proposed model predictive control system is implemented to control the air-conditioning and mechanical ventilation systems in two single-zone testbeds, an office and a lecture theatre, located in Singapore for experimental evaluation of its control performance. The model predictive control system is compared against the original reactive control system (thermostat in the office and building management system in the lecture theatre) in each testbed. The model predictive control system reduces 58.5% cooling thermal energy consumption in the office and 36.7% cooling electricity consumption in the lecture theatre, as compared to their respective original control. Meanwhile, the indoor thermal comfort in both testbeds is also greatly improved by the model predictive control system. Developing a model predictive control system using machine-learning-based building dynamics models could largely cut down the model construction time to days as compared to its counterpart using physics-based models, which usually take months to construct. However, the machine-learning-based modelling approach could be challenged by lack of building operational data necessary for model training in case of model predictive control development before the building has become operational.
energy & fuels,engineering, chemical
What problem does this paper attempt to address?
The problem that this paper attempts to solve is to optimize building energy efficiency and indoor thermal comfort in building automation and control (BAC) systems through model predictive control (MPC) using adaptive machine learning (ML) models. Specifically, researchers are concerned with how to use MPC in combination with ML techniques to overcome the application limitations of traditional physical - model - based MPC in large and complex buildings, such as long modeling time and the need for a large amount of professional knowledge, while improving the energy efficiency of the building and the comfort of its occupants. The main contributions of the paper are as follows: 1. **Developed an adaptive ML building model**: This model can be automatically updated according to online building operation data, so as to better adapt to the changes of the building over time, such as changes in external weather conditions or internal furniture layout. 2. **Proposed a hybrid optimizer**: It is used to solve nonlinear multi - objective optimization problems. This optimizer combines the advantages of global optimization methods and the gradient descent method, improving the solution speed and accuracy. 3. **Conducted long - term experimental evaluations**: The proposed MPC system was implemented in two real buildings (an office and a lecture hall) at Nanyang Technological University in Singapore and compared with the original PID - based BMS system to verify the performance of the system. The research results show that, compared with traditional control methods, this MPC system achieved 36% - 58% cooling energy savings in the two test buildings respectively, while significantly improving the indoor thermal comfort. This proves the effectiveness and potential of MPC using adaptive ML models in optimizing building energy efficiency and comfort.