Minimization of energy consumption in HVAC systems with data-driven models and an interior-point method

• Published in: Energy conversion and management Vol. 85; pp. 146 - 153
• Main Authors: Kusiak, Andrew, Xu, Guanglin, Zhang, Zijun
• Format: Journal Article
• Language: English
• Published: Kidlington Elsevier Ltd 01.09.2014; Elsevier
• Subjects: air conditioning; Air conditioning. Ventilation; Algorithms; Applied sciences; Buildings; Computer simulation; Construction; Energy; Energy consumption; Energy. Thermal use of fuels; Exact sciences and technology; heat; Heating; Heating, air conditioning and ventilation; HVAC; Interior-point method; Internal heat gain; Mathematical models; Multilayer perceptron; neural networks; Nonlinear optimization model; Optimization; Poisson process; simulation models; stochastic processes; Techniques, equipment. Control. Metering; Time-series method; uncertainty; HVAC; Internal heat gain; Multilayer perceptron; Time-series method; Nonlinear optimization model; Poisson process; Interior-point method; Air conditioning; Energy consumption; Time series; Minimization; Modeling; Optimization; Vertebrata; Pisces; Heat gain; Models
• DOI: 10.1016/j.enconman.2014.05.053
• ISSN: 0196-8904

•We study the energy saving of HVAC systems with a data-driven approach.•We conduct an in-depth analysis of the topology of developed Neural Network based HVAC model.•We apply interior-point method to solving a Neural Network based HVAC optimization model.•The uncertain building occupancy is incorporated in the minimization of HVAC energy consumption.•A significant potential of saving HVAC energy is discovered. In this paper, a data-driven approach is applied to minimize energy consumption of a heating, ventilating, and air conditioning (HVAC) system while maintaining the thermal comfort of a building with uncertain occupancy level. The uncertainty of arrival and departure rate of occupants is modeled by the Poisson and uniform distributions, respectively. The internal heating gain is calculated from the stochastic process of the building occupancy. Based on the observed and simulated data, a multilayer perceptron algorithm is employed to model and simulate the HVAC system. The data-driven models accurately predict future performance of the HVAC system based on the control settings and the observed historical information. An optimization model is formulated and solved with the interior-point method. The optimization results are compared with the results produced by the simulation models.