Predictive control of radiant floor air conditioning systems based on an artificial neural network

Due to the large thermal capacity of floor, it is necessary to consider the influence of indoor thermal lag in control of radiant floor air conditioning systems. Develops a predictive control model for the radiant floor air conditioning system with heat pump as cold and heat sources based on RBF artificial neural network. The model can control run time of heat pump and regulate indoor temperature according to the predicted indoor temperature at the next time. Applies the model to regulate the indoor temperature of an experimental radiant floor air conditioning system in heating period. The result shows that the predicted values of indoor temperature are very consistent with real values.