Reserch On The Optimal Confidence Of Wind Power

With the large-scale integration of wind power, unit commitment and economic dispatch have to take the uncertainty of wind power into consideration, which can be described by the confidence intervals. In the present study, the determination of confidence intervals is usually determined by experience, lacking theory basis. However, different confidence intervals correspond to different operating costs and risk costs. The higher the confidence is, the larger the forecast interval of wind power is. At this time, the operating cost is higher to accommodate the uncertainty of wind power within the confidence interval. Due to the large interval of wind power, the probability of wind power without the confidence interval is low, which means the risk and the risk cost are low. On the contrary, the lower the confidence is, the smaller the forecast interval of wind power is. At this time, the risk cost is higher while the operating cost is lower. Thus, the operating cost becomes higher while the risk cost becomes lower. Based on the concept of the confidence of wind power, the overall cost of the system under the confidence of wind power can be calculated. Simulation results show whether the optimal confidence of wind power exists to minimize the overall cost, which can help further research on interval determination of unit commitment. This paper discusses the definition of the optimal confidence of wind power, and proposes a determination method of optimal confidence of wind power. The main work and the obtained results are as follows:1. The concept of optimal confidence of wind power is proposed based on the model of traditional economic dispatch, and the model of optimal confidence of wind power is discussed in detail.2. The determination method of optimal confidence of wind power under simplified cases is proposed. The cases are meant to satisfy the assumptions as below:the forecast error of is normal distributed; the values of optimal confidence of wind power at different time are the same; the forecast interval is symmetrical about the forecast curve of wind power. Based on the assumptions, the output interval can be uniquely determined. Thus, the overall cost can be figured out under different confidence intervals, that can help to establish a relationship between the overall cost and the confidence of wind power, and the model of optimal confidence of wind power can be simplified. An improved particle swarm algorithm is proposed to solve the problem, which accelerates the convergence rate of the swarms. Results shows that, the overall cost first decreases and then raises with the reduction of the confidence of wind power, which proves the existence of the optimal confidence. At the same time, the wind power optimal economic integration interval can be determined by the confidence of wind power, which helps to decide the unit commitment of economic dispatch.3. The determination method of optimal confidence of wind power under complicated cases is proposed. The cases can be classified into two parts:the values of optimal confidence of wind power at different time are different while the forecast interval is symmetrical about the forecast curve of wind power; the values of optimal confidence of wind power at different time are the same while the forecast interval is unsymmetrical about the forecast curve of wind power. Results indicate that, the optimal confidence is existed to minimize the overall cost. The confidence and the wind power optimal economic integration interval are determined by the actual output of the units. Finally, the determination method of optimal confidence of wind power is presented when the values of optimal confidence of wind power at different time are different while the forecast interval is unsymmetrical about the forecast curve of wind power.4. Different factors are discussed related to the optimal confidence of wind power. The factors includes different distributions of forecast error and the correlation between the load and the wind power.