基于模块统一脉宽调制的模块化多电平变换器优化模型预测控制

发布时间：2018-05-19 16:55

本文选题：模块化多电平变换器 + 模型预测控制　；参考：《中国电机工程学报》2017年05期

【摘要】：针对模块化多电平变换器传统模型预测算法(model predictive control,MPC)计算量大、权重因子调节复杂的问题,提出了一种优化的模型预测控制策略。通过控制各相上、下桥臂导通子模块的个数,控制交流侧电流快速跟随给定值,并抑制模块化多电平变换器(modular multilevel converter,MMC)内部环流。同时,基于子模块电容电压排序算法,稳定各子模块电容电压。相较于传统的MMC模型预测控制方法,该方法无需求解价值函数,很大程度上降低了控制器的运算量,同时有效控制了相电流、环流与子模块电容电压,适用于控制性能要求较高的场合。借助模块化多电平逆变器仿真模型与实验样机验证所提算法的有效性。结果表明,优化后的模型预测算法能有效抑制环流,维持子模块电容电压稳定,相电流快速跟随给定值,同时很大程度上减少了控制系统的计算量。
[Abstract]:A model predictive control strategy is proposed for modularized multilevel converters, which is characterized by large computational complexity and complex weight factor adjustment. By controlling the number of the lower arm conduction modules in each phase, the AC side current is controlled to follow the given value quickly, and the internal circulation of the modularized multilevel converter is suppressed. At the same time, the capacitor voltage of each submodule is stabilized based on the capacitor voltage sort algorithm of the submodule. Compared with the traditional predictive control method of MMC model, this method does not need to solve the value function, greatly reduces the computation of the controller, and effectively controls the phase current, the circulation and the capacitor voltage of the sub-module. It is suitable for the situation with high control performance. The effectiveness of the proposed algorithm is verified by modularized multilevel inverter simulation model and experimental prototype. The results show that the optimized model prediction algorithm can effectively suppress the circulation, maintain the capacitance voltage stability of the submodule, and rapidly follow the given value of the phase current, at the same time, greatly reduce the calculation of the control system.
【作者单位】：中国矿业大学信息与电气工程学院;
【基金】：国家自然科学基金项目(51377160) 江苏青蓝工程资助项目(04150020)~~
【分类号】：TM46

【相似文献】