Research On Gas Multi-period Dispatch Based On Gasholder Operation Optimization

As an important secondary energy source for steel companies,by-product gas accounts for about 30%of the company’s total energy.By making full use of this part of energy,companies can reduce the operating costs of the gas system and increase their profits.Through reasonable energy distribution,energy consumption and carbon emissions can be reduced,and enterprise energy conservation and emission reduction goals can be achieved.In addition to the main process equipment,large steel companies often also set up buffer equipment for buffering gas volume and pressure fluctuations,such as boiler equipment,hot-rolling mixing stations and gasholders.In previous studies,most of the gasholders were used as security equipment and failed to pay attention to its important role in the gas system.In this paper,the gasholder is regarded as the main adjustment equipment of the supply and demand balance of the gas system,and a gas multi-period dispatch model based on the optimization of the gasholder operation is established.In order to improve the accuracy of optimal gas dispatch,this paper establishes a gas dynamic prediction model.By using the gas supply predicted by the model as the input variable of the gas multi-period dispatch model,the optimal dispatch in the next 8 hours can be achieved,and the optimization effect is good.Through investigating the actual production data of iron and steel enterprises,the gas generation and consumption characteristics are obtained.Based on the popular XGBoost algorithm in machine learning,gas dynamic prediction models under different working conditions are established.Two different working conditions are set in the paper,namely normal production conditions and blast furnace maintenance conditions.The model predicts the gas generation and consumption of the two operating conditions in turn,and calculates the prediction error for the equipment with large fluctuations.The results show that the prediction error of blast furnace generation under normal production conditions is 1.16%,and the prediction error of hot blast stove consumption BFG is 3.65%;the prediction error of blast furnace generation under blast furnace maintenance conditions is 2.59%,and the prediction error of hot blast stove consumption BFG is 4.42%.A gas multi-period dispatching model was established,with the objective function of minimizing the operating cost,and the time-of-use electricity price was introduced into the model.Based on the gas dynamic forecast,the allocation of the gas supply among the buffer equipment was optimized.Scenarios are set for different gas supply and different time periods of dispatch.The results show that the model can allocate the supply gas more reasonably,and realizes the strategy of purchasing less electricity and generating more electricity in the peak electricity price stage,and the strategy of buying more electricity and generating less electricity in the valley electricity price stage can effectively reduce the cost of 10%～20%electricity purchase,greatly saving production and operation costs.