A systematic aeration control strategy aiming at energy saving was proposed using a comprehensive model composing of a mass flow analysis (MFA) module, an oxygen transfer rate (OTR) module and an aeration optimization module. The model was calibrated and validated using data sets from a full-scale anaerobic/anoxic/oxic (AAO) process wastewater treatment plant (WWTP). The MFA module was extended with a detailed description of total oxygen demand (TOD). The oxygen supply was predicted considering the negative correlation between soluble COD and α factor, which quantified the dynamic effect of wastewater contaminants on aeration equipment. Energy saving potential (ESP) was assessed via model-predictive optimum air supply. After the implementation of the aeration control strategy, the ESP of the blowers was reduced from 26.3 % to 12.2 %, without worsening the overall pollutant removal. As the aeration system can respond in advance to the varying influent characteristics, model-predictive control together with sensor-based control holds promise for application in full-scale WWTPs for energy conservation.