【摘要】 以需求响应为核心进行配电网负荷控制优化时，节能百分比较低。因此，提出基于高斯过程回归的主动配电网负荷控制优化方法。依托于随机森林算法，判断所有输入变量的重要性，明确主动配电网负荷影响变量。运用高斯过程回归原理，应用负荷预测算法，准确得出主动配电网需求侧未来一段时间内的负荷要求。在此基础上，以直接负荷控制为核心，建立负荷控制协调优化模型，再应用遗传算法求解最优负荷控制优化方案。算例分析结果表明：所提优化方法的节能百分比达到6.23%，说明该方法达到了节能的目的。更多还原

【Abstract】 When the distribution network load control optimization is based on demand response, the energy saving percentage is relatively low. Therefore, an active distribution network load control optimization method based on Gaussian process regression is proposed. Based on the random forest algorithm, the importance of all input variables is judged, and the load influence variables of the active distribution network are determined. Based on the principle of Gauss process regression and load forecasting algorithm,the load demand of active distribution network in the future can be accurately obtained. On this basis, the load control coordination optimization model is established with the direct load control as the core, and then the genetic algorithm is applied to solve the optimal load control optimization scheme. The analysis result of the example shows that the energy saving percentage of the proposed optimization method reaches 6.23%, which indicates that the method achieves the purpose of energy saving.更多还原