The “dual carbon” goals will promote the continuous application of wind power and other renewable energy. With the large-scale integration of wind power, there would be some risks when the power system operation because of the inherent uncertainty of wind power. While, the traditional deterministic method does not consider the wind power prediction error, and the unit reserves enough reserve capacity to deal with the uncertainty of wind power, so the system has great hidden Security Problems. In recent years, many scholars have constructed robust optimization models based on wind power prediction errors, but the results tend to be conservative. To address these issues, this paper proposes a bi-level optimization model which considering times series characteristic of wind power forecast error and wind power reliability. It effectively improves the economy of power system operation. Firstly, the adaptive bandwidth method is used to obtain the non-parametric kernel density estimation function of the prediction error, and the time series segment of wind power prediction error is optimized through correlation analysis, and the fluctuation domain of wind power is established according to the time series segment, and the intra-day wind power scenario is generated. Secondly, the bi-level optimization model is constructed. The upper model in the day-ahead phase the objective function is to maximize the utilization of wind power and minimize the generation cost and carbon transaction cost, to solve the planned output of each unit, wind power and allowable output area of wind power. The planned output of wind power is determined according to the reliability of intra-day wind power scenario. The allowable output area of wind power makes the control of wind power plant more flexible, and determines the output decision of Automatic Generation Control(AGC) units through participation factors to deal with wind power fluctuations. While the lower model in the intraday generates wind power scenarios take the system deviation correction cost and risk cost minimization as the objective function, the source-side considers Ns possible scenarios to get the reliability of wind power in stages and feedback to the upper model, the incentive demand response is introduced on the load side, and the lower model updated the allowable output area of wind power and adjusts the output of AGC units by tracking the planned output value obtained from the upper model. Finally, the proposed model is compared with other models based on the data of a certain region in Xinjiang, and the results are analyzed. A total of five scenario models are compared. The results show that in scenario 1, the cost is the lowest because the uncertainty of wind power is not considered; in scenario 2, the unit commitment result is conservative lead the cost highest; in scenario 4, the cost is higher than scenario 3 presented because does not distinguish AGC units, and all thermal power units track the command value of the plan and reserve enough spare capacity. In Scenario 5, the set of the same participation factor lead to same priority among AGC units, and distribute power equally to each unit during unit operation, so the cost increases compared with scenario 3. thermal interaction, can reduce equipment capacity waste, achieve higher overall benefits. The following conclusions can be drawn from the simulation analysis: (1) The distributed "point-middle-center" storage and transportation mode is adopted in the multi-regional system, which can be transported by the adjacent temporary storage station according to the demand of straw fuel in each park, thus saving the transportation cost. (2) The multi-regional collaborative planning scheme based on biomass "point-middle-center" storage and transportation mode by transporting raw materials can optimize the allocation of straw resources, improve the utilization ratio of equipment, reduce capacity waste, and improve the overall economy. (3) From the aspects of biomass storage and transportation, energy supply mode and system coordination planning model, the method presented in this paper and the above conclusions are universal. © 2023 Chinese Machine Press. All rights reserved.
