Using Model Predictive Control to Modulate the Humidity in a Broiler House and Effect on Energy Consumption

In moderate climate, broiler chicken houses are important heating energy consumers and hence heating fuel consumption accounts for a large part in operating costs. They can be reduced by constructional measures, which in turn lead to important costs as well. On the other hand, a software solution to reduce energy would lead to considerably less follow-up costs. The main objective of our work was to assess if it is possible to save energy with a software solution and eventually quantify the savings for a given broiler house in the Swiss Plateau. The investigation was carried out in simulation: the particular broiler house was measured, and a dynamical model for it was derived and validated. To actually search for a particular behaviour of the software that would lead to energy savings, model predictive control was used. The idea was not to specify a particular behaviour of the software but rather to let the software itself find the best behaviour in an exhaustive search. The simulations showed that energy savings can be realised mainly by letting the indoor humidity deviate from what usually is used as setpoint and hence take profit of the outdoor climate, which changes naturally during a 24-hour course. We used expert opinions to determine how long and large these setpoint deviations may be without harming the broilers. The simulations showed alsothat the light control and the biological activity of the animals reduced the potential savings.