文献类型：期刊文献

中文题名：基于模型预测的微电网日内分层调度研究

英文题名：Research on Intra-day Hierarchical Dispatching of Microgrid Based on Model Prediction

作者：张恒 张靖 肖迎群 何宇 刘影 范璐钦

第一作者：张恒

机构：[1]贵州大学电气工程学院,贵州贵阳550025;[2]贵州理工学院大数据学院,贵州贵阳550003;[3]贵州电网公司电网规划研究中心,贵州贵阳550001

第一机构：贵州大学电气工程学院,贵州贵阳550025

年份：2021

卷号：37

期号：4

起止页码：1-10

中文期刊名：电力科学与工程

外文期刊名：Electric Power Science and Engineering

基金：国家自然科学基金(51867005);贵州省科技计划项目([2018]5615);贵州省科技计划项目([2018]5781)。

语种：中文

中文关键词：不确定因素;分层模型预测控制;日内协调运行;误差反馈;改进鲸鱼优化算法

外文关键词：uncertain factors;hierarchical model predictive control;intra-day coordinated operation;error feedback;improved whale optimization algorithm

摘要：随着可再生能源的大规模接入,可再生能源和负荷不确定因素给微电网优化调度带来了新的挑战。提出基于模型预测的微电网分层调度模型,采用分层模型预测控制策略实现“源–网–荷–储”的日内协调运行。日内优化层以运行成本最小为目标函数,为日内修正层提供参考值。日内修正层以修正后的调度计划与参考值偏差最小为目标函数,并根据预测输出与实测输出间的误差构成反馈环节,以解决不确定因素对优化调度的影响。同时,为改善计算速度和精度,在模型预测控制中引入改进鲸鱼优化算法。算例结果表明,基于模型预测的微电网日内分层调度策略具有较好的可行性和有效性。
With the large-scale integration of renewable energy,renewable energy and load uncertainties have brought huge challenges to the optimal dispatching of microgrid.This paper proposes a hierarchical dispatching model for microgrid based on model prediction,and adopts predictive control strategies of the hierarchical model to realize the intra-day coordinated operation of“generation-grid-load-storage”.The intra-day optimization layer takes the minimum operating cost as the objective function to provide reference values for the intra-day correction layer;the intra-day correction layer takes the minimum deviation between the revised scheduling plan and the reference value as the objective function,and forms a feedback based on the error between the predictive output and the actual measured output to solve the influence of uncertain factors on optimal scheduling.At the same time,in order to improve the calculation speed and accuracy,an improved whale optimization algorithm is introduced into the model predictive control.The results of calculation examples show that the intra-day hierarchical dispatching strategy of microgrid based on model prediction is feasible and effective.