Search

Model predictive control (MPC) is a recent development that uses modelling and simulation with forecast conditions to exploit building thermal mass in an effort to improve indoor thermal comfort and reduce energy use/cost compared with traditional rule-based control strategies. This paper investigates a range of recent MPC strategies and the direction in which this new technology is trending. The findings indicate that there is little evidence that directly compares the performance of specific optimization algorithms, forecast and simulation parameters (timestep, horizon), and climate forecast accuracy for the same scenario. An analysis of 19 case studies highlights the advantages of MPC compared with conventional control strategies, but also identifies areas that need further improvement. These areas include the optimization strategy, the effects of forecast disturbance assessment, and desirable traits of existing buildings under consideration for MPC implementation. This paper develops a set of target parameters for the types of buildings for which MPC will have the most impact and suggests methods for overcoming shortfalls identified in recently published studies. [ABSTRACT FROM AUTHOR]