引入参数辨识的并网变换器事件触发预测控制  

Event-triggered Predictive Control of Grid-connected Converter With Parameter Identification

作　　者：徐瑞芬[1] 叶钢[1] 刘佳[2] 曹频[2,3] XU Rui-fen;YE Gang;LIU Jia;CAO Pin(Lishui Vocational and Technical College,Lishui 323000,China;不详)

机构地区：[1]丽水职业技术学院,电子信息工程系,浙江丽水323000 [2]浙江大学,电气工程系,浙江杭州310058 [3]杭州晶耐科光电技术有限公司,浙江杭州310058

出　　处：《电力电子技术》2024年第8期99-102,共4页Power Electronics

基　　金：浙江省教育厅一般科研项目(Y202147936);杭州晶耐科光电技术有限公司委托项目(EY202109)。

摘　　要：经典的三电平并网变换器模型预测算法具有计算资源占用度高、开关损耗大、控制精度受模型参数影响较大等缺点。为提高预测控制精度并降低系统损耗,此处提出一种引入参数辨识的三电平并网变换器的事件触发型有限控制集模型预测算法,该算法通过构建系统状态和动态事件触发条件之间的解析表达方程,利用系统状态实时反馈设置跟踪电流误差阈值的事件触发条件以减少系统计算损耗,同时,引入参数在线辨识,有效提高了系统参数鲁棒性,并实现了在尽量保持较小计算量的情况下还具有改善并网电流质量的优点。最后,构建了一台三电平T型并网变换器平台,实验结果表明,此处所提算法较之经典的模型预测方法有较大改善,具有较大的应用价值。The classical three-level grid-connected converter model prediction algorithm has some disadvantages,such as high computing resource consumption,large switching loss,and large control accuracy influenced by model parame-ters.In order to improve predictive control accuracy and reduce system loss,an event-triggered finite control set model prediction algorithm of three-level grid-connected converter with parameter identification is proposed.By constructing the analytic expression equation between the system state and the dynamic event trigger condition,the algorithm uses the real-time feedback of the system state to set the event trigger condition of tracking current error threshold to re-duce the calculation loss of the system.At the same time,the parameter online identification is introduced to effective-ly improve the robustness of the system parameters.And it has the advantage of improving the quality of grid-connect-ed current while minimizing computational complexity.Finally,by constructing a three-level T-type grid-connected converter platform,the experimental results demonstrate significant improvement compared to classical model predictive methods,highlighting its considerable practical value.

关 键 词：三电平并网变换器 模型预测 参数辨识 事件触发 

分 类 号：TM46[电气工程—电器]