基于灵活性矩阵与优劣解距离的新能源功率预测评价方法

An Evaluation Method for New Energy Power Prediction Based on Flexibility Matrix and TOPSIS Method

鉴于功率预测评价对电力系统运行和调度的重要性,文章充分考虑电网调度实际需求,从不同时期下机组的出力大小和不同时段下电网调度对预测精度的需求2个方面分析现有功率预测评价指标的不合理之处,并提出了新的准确率指标。首先,对不同时期下系统灵活性进行计算得到灵活性矩阵并对其进行预处理;其次,将处理后的灵活性矩阵输入到优劣解距离法(technique for order preference by similarity to an ideal solution,TOPSIS)模型中,根据评价对象与理想化目标的接近程度对不同时段的灵活性进行排序和归一化计算,得到不同时段下的权重系数矩阵,进而形成新的评价指标;最后,将所提的评价指标与其他指标进行对比。仿真结果表明,所提评价指标更能体现系统灵活性对功率预测评价的影响,满足实际调度的需求。

Considering the importance of evaluation of power prediction for power system operation and dispatching, this paper takes the actual demand of gird dispatching into consideration and analyzes the irrationality of existing prediction evaluation indices from two aspects: the output of units in different periods and the demand of grid dispatching for forecasting accuracy in different periods. And some new evaluation indices are put forward. Firstly, the flexibility matrix of the system in different periods is calculated and pretreated. Secondly, the processed flexibility matrix is input into the technique for order preference by similarity to an ideal solution(TOPSIS) model. According to the closeness of the evaluation object and the idealized target, the flexibility in different periods is sorted and normalized to obtain the weight coefficient matrix, and then a new evaluation index is formed. Finally, the evaluation index proposed in this paper is compared with other indices. The simulation results show that the evaluation indices proposed in this paper can better reflect the influence of system flexibility on power prediction evaluation and meet the requirements of grid dispatching.