风光储微网系统多时间尺度模型预测调度策略

Model Predictive Control Based Multiple-Time-Scheduling Strategy for Wind-pv-es Hybrid System

针对风光储微网系统,建立基于模型预测的多时间尺度微网调度策略。首先,为提高微网调度运行准确性,采用灰色GM(1,N)-BP神经网络方法构建微网发电量及负荷需求的预测模型,然后构建基于预测模型的多时间尺度微网调度优化策略,该策略考虑了储能装置充放电运行的功率变化,以日前调度中的能源出力为参考,在日内调度阶段采用滚动修正,通过细化滚动时间,实现微网运行调度的精确修正,减少储能装置的充放电损耗。最后,基于微网运行数据进行了算例分析,验证预测模型与调度策略的有效性。结果表明,相较于传统单一灰色预测方法,所采用的组合预测模型预测精度更高,所构建的多时间尺度调度策略使得清洁能源得到充分利用,实现微网经济安全运行。

A model predictive control based multiple time scheduling strategy is proposed for the wind-pv-energy storage hybrid system.Fistly,a power generation and load demand forecasting model based on gray GM(1,N)-BP neural network is optimized to improve the accuracy of microgrid dispatch operation.Then,a model predictive control based multiple time scheduling strategy is proposed,this strategy giving full consideration to battery power change during charge and discharge operation,takes the day-ahead dispatch plan as the reference,establishes the prediction model based on the multi-step rolling optimization to achieve accurate correction of microgrid operation scheduling,reduce charge and discharge losses of energy storage devices.Finally,the simulation experiment analysis on the proposed control strategy is carried out in order to verify its effectiveness.It is proved that the predictive model better than the traditional single gray prediction strategy,the multiple time scheduling strategy is able to make full use of clean energy,and ensures the systerh running with high economic eficiency and security.