Study On Optimal Peak-shaving Method For Thermoelectric Units Considering The Confidence Of Wind Power Prediction

With the coming of the electric power Internet of things era,the installed proportion of new energy is increasing,and the wind abandonment caused by the problems of wind power generation as the representative of new energy,its volatility and intermittence is becoming more and more serious.Because of its winter heating period in the north of China,the phenomenon of winter wind abandonment in the north is more prominent due to the operation mechanism of the thermoelectric unit "setting electricity by heat ".During the winter heating period,the electric output of large thermal power units in the north is restricted by the heat load,so it is difficult to provide more output space for wind power absorption.In view of the above phenomena,this paper first studies the wind power fluctuation characteristics in the northern part of China,and analyzes the fluctuation characteristics of wind power generation in winter heating period from the operating mechanism of wind power generation.On this basis,the wind power prediction model and prediction error are analyzed,and the uncertain model of wind power output is established.Based on the prediction error of wind power,the reliability function of wind power prediction defined in this paper.Determine the confidence interval for wind power output in the future.The optimal dispatching method of wind turbine considering the reliability of wind power prediction is put forward in turn.In this paper,adding wind power to the conventional power supply balance can reduce the rotary reserve and greatly reduce the economy of power system operation.Based on the above research,this paper deeply analyzes the thermoelectric characteristics of thermoelectric units.The relationship between the electric and thermal output of the thermoelectric unit and the effect of heat storage on the electric heating characteristics of the thermoelectric unit are analyzed.Finally,this paper proposes an ultra-short-term control strategy for distribution network considering the reliability of wind power prediction and the characteristics of thermoelectric units,and uses model predictive control to optimize distribution network scheduling problems.by collecting the parameters of the system at the current moment and considering the constraints of the current moment and the future moment,the model achieves the real-time optimal control of the dynamic system through MPC rolling optimization.The control strategy aims at the minimum quantity of abandoned wind power,which can increase or reduce the power output in a short time by regulating the thermoelectric unit,and adjust the heat load demand of the system through the heat storage device.The electric heating load distribution of the system in each control period is considered to maximize the wind power output.and the effectiveness of the proposed method is verified by simulation analysis.