Research On Energy-saving Application Of Pumping Station Based On Machine Learning Algorithm

Wastewater treatment plant is an important method to solve the water pollution problem.With the increase of water consumption per capita,the water pollution problem is becoming more and more serious,and there are more and more wastewater treatment plants in recent years,but due to the excessive energy consumption of wastewater treatment plants,resulting in many wastewater treatment plants are facing the problem of insufficient cost costs.In wastewater treatment plants,pump and blower systems account for more than 80% of the energy costs of wastewater treatment plants.Energy efficiency studies of equipment in wastewater treatment plants can improve the energy efficiency of wastewater treatment and reduce the operating costs of wastewater treatment plants.In this paper,we optimize pumping station energy consumption by improving the operation schedule of the lift pumping station system in the wastewater treatment process,and establish a cluster class optimization algorithm based on the energy-saving operation rules of the lift pumping station.Based on the operation logic of the pumping station system in wastewater treatment plants,a pumping station-related model based on machine learning is established.The GRU network was used to model the wastewater connection pond influent of the wastewater treatment plant,and the model was used to predict the influent of the wastewater connection pond and calculate the required pumping capacity of the pumping station in the future by using certain calculation rules.The energy consumption and flow rate of the lift pumping station were modeled using the GPR algorithm,and the energy consumption and pumping volume of the pumping station operating condition were predicted using the energy consumption model and flow rate model.A model for optimizing the operation of a lift pumping station using the GRU network water intake prediction model as the operating constraint and the GPR energy consumption and pumping volume prediction model as the evaluation function was developed,which uses an improved particle swarm optimization algorithm to minimize the energy consumption of the lift pumping station and ensure the pumping volume of the lift pumping station as the goal.The model was validated with high accuracy for the GRU wastewater connection tank influent prediction,GPR energy consumption prediction model,and GPR flow prediction model using a Northeast wastewater treatment plant as a case study.The energy consumption optimization model was verified,and energy savings of about 10% were achieved compared to the energy consumption of conventional automation.