ANN-based sensorless adaptive temperature control system to improve methane yield in an anaerobic digester

Constant methane yield from the biogas plants is necessary to achieve stable heat and power generation. From conventional biogas plants, a fluctuating methane yield is obtained due to variation in operating conditions. In this paper, an inferential control for constant methane yield by regulating the digester temperature is proposed. The optimal operating temperature of the digester is determined using artificial neural network (ANN). It considers variations in total volatile solids and hydraulic retention period to get constant methane yield. After training the proposed ANN, achieved MSE is 0.0003522, RMSE is 0.01876, and R-2 of 1 for training, validation, and testing. A proportional-integral-derivative controller tuned by bacterial foraging-particle swarm optimization along with a derivative filter has been used in the temperature control loop. In addition, the temperature sensor is replaced by a temperature estimator in the control loop. The performance of the proposed control scheme has been examined for various realistic operating conditions using MATLAB software.