Exploration of deep learning models for real-time monitoring of state and performance of anaerobic digestion with online sensors

•Sensor data of pH, EC, and ORP interrelate with the state and performance of AD.•CNN extracts features from sensor data, and LSTM infers output variables.•A combined model of CNN and LSTM predicts almost all AD variables from sensor data.•The combined model is an excellent soft sensor for the real-time monitoring of AD.•The combined model helps the stable operation of AD through real-time monitoring.