基于改进模型预测控制的“大机小网”下火电-光伏-抽水蓄能优化调度

Optimization Operation of Thermal-Photovoltaic-Pumped Storage in“Large Machine and Small Network”Based on Improved Model Predictive Control

光伏大规模接入加剧了"大机小网"电网的结构性问题,导致传统单断面优化调度难以实现最优运行。提出了一种基于改进模型预测控制(MPC)的多时间尺度火电-光伏-抽水蓄能联合优化调度模型。即对于日前优化调度,在系统运行成本中引入了光伏未消纳量和火电污染物排放量形成多目标函数;对于日内滚动优化,以跟踪日前计划值为目标,考虑了光伏出力极限场景调节能力约束与抽水蓄能机组启停速率约束,并采用自适应调节预测时域与控制时域的MPC方法改进了传统MPC。海南省某区域电网的实际运行验证了该综合模型的有效性。

Large-scale photovoltaic power connected to the grid intensifies the structural problem of"large machine and small network".It is difficult for traditional single-section optimal dispatching to achieve optimal operation.A multitime scale model of optimization operation for thermal plant,photovoltaic generation and pumped storage station is proposed based on improved model predictive control(MPC).For the day-ahead optimal dispatch,a multi-objective function is constructed through integrating the system's operation cost with photovoltaic power curtailment and thermal power emission of pollutants.For intra-day rolling optimization dispatch,its goal is to track the day-ahead schedule value.The regulation capacity in photovoltaic output limit scene and start-stop rate of pumped storage units are considered as the constraints,and the traditional MPC is improved by adaptive regulation of predictive and control domains.The validity of the integrative model is verified by an example of a regional power grid in Hainan Province.