A multi-layer perceptron neural network is applied to construct the optimization
model. Due to high computational complexity of the model, a particle swarm optimization algorithm is used to maximize the biogas production by finding the optimal settings of controllable variables.