计及功率预测误差的主动配电网运行方式优化-《郑州大学学报(工学版)》

文章信息/Info

Title:: Operation mode optimization of active distribution network considering power prediction error

Author(s):: Wang Kewen; Liu Kai; Liu Yanhong; School of Electrical Engineering, Zhengzhou University

Keywords:: active distribution networksecond order correctionprobability power flow trust regioncon-straint condition

摘要:: 在主动配电网的运行方式优化中，节点功率数据通常来自功率预测，存在预测误差及相应的分布特性，从而可以采用概率表达进行描述。计及功率预测误差的分布特征，以综合运行费用均值为目标函数，节点功率平衡方程为等式约束，节点电压和支路功率等变量的运行范围构成不等式约束，建立主动配电网运行方式的概率优化模型。通过分析优化算式的特点，采用二阶潮流表达的概率描述，在随机变量的均值计算中计及方差修正，提高均值计算的准确度。在优化模型求解中，依据变量的实际特点，对离散变量和连续变量采用不同的处理方式，应用信赖域管理技术处理连续变量。通过118节点算例的计算分析，说明所述算法的可行性和实用性。

Abstract:: In the optimization of the operation mode of active distribution network, node power data usually comes from power prediction, and there are prediction errors and corresponding distribution characteristics, which can be described by probabilistic expression. Considering the distribution characteristics of power prediction error, taking the average of comprehensive operating costs as the objective function, the node power balance equation as an equality constraint, and the operating range of variables such as node voltage and branch power constituting inequality constraints, the operating mode of active distribution network is established. Probabilistic optimization model. By analyzing the characteristics of the optimization formula, the probability description of the second-order power flow expression is adopted, and the variance correction is taken into account in the calculation of the mean value of random variables to improve the accuracy of the mean value calculation. In solving the optimization model, according to the actual characteristics of variables, different processing methods are used for discrete variables and continuous variables, and trust region management technology is used to process continuous variables. The feasibility and practicability of the algorithm are illustrated through the calculation and analysis of a 118-node example.