基于超短期风电功率预测的混合储能控制策略研究

发布时间：2018-11-06 15:04

【摘要】：为了改善风机出力特性,提出了一种基于超短期风电功率预测的混合储能控制策略。首先,利用解析模态分解方法从风电信号中提取低频信号,采用了一种改进布谷鸟方法优化支持向量机的惩罚因子参数和核函数参数进行超短期功率预测;然后,对低频预测信号建立1 min时间尺度和30 min时间尺度的功率波动并网指标,判断是否触发蓄电池动作,若动作,采用AMD分解自适应调整低频预测信号的截止频率,直到满足并网要求,确定蓄电池补偿功率指令。最后根据蓄电池荷电状态和补偿功率指令自适应调节原始风电信号截止频率,高频信号通过模糊控制由超级电容器补偿。仿真算例表明,该方法可以有效平滑功率波动,减少蓄电池的循环次数,同时保证了蓄电池储能的平滑能力,避免过充过放,延长蓄电池的寿命。
[Abstract]:In order to improve the performance of blower, a hybrid energy storage control strategy based on ultra-short term wind power prediction is proposed. Firstly, the low frequency signal is extracted from the wind signal by the analytic mode decomposition method, and an improved Cuckoo method is used to optimize the penalty factor parameter and kernel function parameter of support vector machine for ultra-short term power prediction. Then, the power fluctuation indexes of 1 min time scale and 30 min time scale are established for the low frequency prediction signal to judge whether or not to trigger the battery action. If the operation is done, the cut-off frequency of the low frequency prediction signal is adjusted adaptively by AMD decomposition. Until the grid connection requirements are met, battery compensation power instructions are determined. Finally, the cutoff frequency of the original wind signal is adaptively adjusted according to the charging state of the battery and the compensation power instruction, and the high frequency signal is compensated by the supercapacitor through fuzzy control. The simulation results show that the proposed method can effectively smooth the power fluctuation, reduce the cycle times of the battery, ensure the smooth energy storage ability of the battery, avoid overcharging and overdischarge, and prolong the battery life.
【作者单位】：华北电力大学河北省输变电设备安全防御重点实验室;
【基金】：国家自然科学基金资助项目(51177047)
【分类号】：TM614

【相似文献】