基于多步模型预测控制的模块化多电平换流器环流控制策略

发布时间：2018-07-02 14:09

本文选题：模块化多电平换流器 + 环流抑制　；参考：《电力系统自动化》2017年16期

【摘要】：模块化多电平换流器(MMC)以其结构模块化、低谐波输出以及冗余控制等优点在中高压领域得到了广泛的关注与研究。推导了MMC离散数学模型,提出了基于多步模型预测控制的MMC环流控制方法,能够实现桥臂环流的多步优化控制并有效地降低预测控制计算量。首先利用环流电流离散状态方程进行单步环流预测,再选取满足单步预测效果的投入模块数进行多步环流预测,最终求解出桥臂投入模块数的优化解,实现环流电流的多步优化控制,从而有效地抑制环流中的谐波电流。所提多步预测控制利用单步预测得到的优化解构建多步预测的有限控制集,可以大幅减少多步预测所需要的循环预测次数,有效地降低控制器的计算量。最后在PSCAD/EMTDC中搭建了201电平MMC时域仿真控制系统,仿真结果验证了所提方法的正确性与有效性。
[Abstract]:Modularized multilevel converter (MMC) has been widely studied in the field of medium and high voltage due to its advantages of modularization, low harmonic output and redundant control. The MMC discrete mathematical model is derived, and the MMC circulation control method based on multi-step model predictive control is proposed, which can realize the multi-step optimal control of the bridge arm circulation and effectively reduce the computational complexity of the predictive control. First, using the discrete state equation of circulation current to predict the single-step circulation, then selecting the number of input modules that satisfy the single-step prediction effect to carry out the multi-step circulation prediction, and finally solving the optimal solution of the input module number of the bridge arm. The multi-step optimal control of the circulation current is realized, thus the harmonic current in the circulation is effectively suppressed. The proposed multi-step predictive control uses the optimization solution obtained by single-step prediction to construct a finite control set of multi-step prediction, which can greatly reduce the cycle prediction times required by multi-step prediction and effectively reduce the computational complexity of the controller. Finally, a 201 level MMC time-domain simulation control system is built in PSCAD / EMTDC. The simulation results verify the correctness and effectiveness of the proposed method.
【作者单位】：国家电能变换与控制工程技术研究中心(湖南大学);
【分类号】：TM46

【相似文献】