Study on low-carbon electricity consumption strategy for photovoltaic power generation wastewater plant based on convolutional neural network

Wastewater treatment plants (WWTPs) are significant electricity consumers, leading to substantial energy use and carbon emissions. In response to China's “dual carbon” goal, photovoltaic power generation systems are increasingly adopted in WWTPs for their economic and environmental advantages. However, photovoltaic power is highly weather-dependent and unstable. This paper proposes a low-carbon electricity consumption strategy for WWTPs with photovoltaic systems by optimizing blower power scheduling to maximize photovoltaic use and reduce electricity costs while ensuring effluent quality. An anaerobic-anoxic-oxic (AAO) WWTP in Wuhan City served as a case study. A water quality simulation model was established using GPS-X, and the photovoltaic system was designed and assessed using PVsyst. Then, the models for water quality, photovoltaic power generation, and electricity pricing were integrated to adjust blower power consumption, forming a set of low-carbon strategy. The strategy sample set was input into the convolutional neural network (CNN) model, which accurately predicted dispatch strategies, achieving a recall and accuracy rates above 85 %. Analysis of the September 2019 operation data indicates that this strategy can yield an additional 10.1 % savings on the WWTP's electricity bill. This approach provides a reference for low-carbon power management in other wastewater plants with photovoltaic systems and promotes effective system integration.