Broiler growth optimization using optimal iterative learning control

In this paper the first recorded attempt at optimizing broiler growth using iterative learning control under state-of-the-art production conditions is presented. The work is motivated by a significant predicted increase in global broiler meat, where existing optimization techniques are incompatible with state-of-the-art broiler production. The proposed method regulates broiler growth using broiler house temperature based on norm optimal iterative learning control, which is a model based control technique. To compensate for the lack of mathematical broiler growth models in scientific literature, dynamic neural network models are used, which is a data driven modeling technique. Practical results from a state-of-the-art broiler house appear promising, but not conclusive, although a maximum decrease in required feed of 2.5% was obtained.