Paper Conference

Proceedings of ASim Conference 2016: 3rd Asia conference of IBPSA-China, Japan, Korea


Hybrid supply air controls using fuzzy inference system and neural network fitting models for control and energy efficiencies

Jonghoon Ahn, Dae Hun, Soolyeon Cho

Abstract: This paper presents hybrid supply air control approaches for control and energy efficiencies, using Fuzzy Inference System (FIS) and Artificial Neural Network (ANN) fitting model. Recently, advanced computing and statistical technologies, such as FIS and ANN algorithm, were introduced to replace the conventional controls for improved energy efficiency in HVAC systems. However, these methods, which were mostly used to control fuel amount or fan motor speed, had lack of the capability of immediate response to the demand of thermal zones. This paper introduces a zone level hybrid control approach. Simultaneous controls of the amount of supply air and its temperature by FIS and ANN algorithms are developed and tested to evaluate the supply air conditions (mass and temperature) for heating season. The sum of errors, caused by the difference between set-point and actual room temperatures is used as an indicator of energy efficiency. The both FIS and ANN models are compared to typical thermostat on/off baseline controller. The results include the total errors of hybrid models in comparison with the baseline controller. This paper analyzes the effectiveness of hybrid controllers using FIS and ANN models, which can be used to optimize mass and temperature of supply air to meet set-point temperature.
Keywords: Hybrid Heating Control, Fuzzy Inference System, Neural Network Fitting, Control Efficiency, Energy Saving