Modeling and control of ventilation and heating system using neuro-fuzzy inference system

Abstract

Dealing with the nonlinearities and uncertainties in Ventilation and Heating System are the main challenges in developing a reliable model for the system. In this project, artificial neural network (ANN) modeling technique was used as it has demonstrated the capability of handling certain uncertainties. The laboratory scale ventilation and heating system, VVS-400 equipped with RTD temperature sensor and orifice plate as flow sensor is chosen as the case study. The input-output data of the system was collected experimentally in building ANN model for the plant. Large portion of the pre-treated data were used to train the ANN model. The remaining portion were used to test the generalization capabilities of the realized ANN model. The prediction performances of the model were evaluated using root-mean square error (RMSE) and correlation coefficient (R). A neuro-fuzzy controller was designed to control the air temperature of the system. The simulation studies were achieved through the use of MATLAB/Simulink software.