摘要
传统的模型预测控制根据当前电流采样和电机数学模型来预测下一时刻的定子电流,对两电平逆变器需要进行7次预测才能得到使定子电流误差最小的电压矢量。其主要不足是对系统模型和电机参数的依赖性较强而且计算量大。本文在分析永磁同步电机电流预测模型的基础上,得到一种只需一次预测和比较即可选出最佳矢量的快速预测电流控制。进一步对其进行简化得到了一种不依赖于电机模型和参数的无模型预测电流控制。该方法在一个控制周期内仅需对定子电流进行单次采样,具有计算量小、鲁棒性好等优点并且动静态性能良好。本文从参数敏感性、稳态和动态性能等方面对快速预测电流控制和无模型预测电流控制进行了对比,仿真和实验结果验证了所提方法的有效性和正确性。
Conventional model predictive control(MPC)predicts the stator current based on thecurrent sampling and mathematical model of motor,which requires seven times of prediction to obtainthe best voltage vector minimizing the current errors for two-level inverter.The main disadvantages aredependence on system model and machine parameters and high computational burden.By analyzing thepredictive model of stator current for permanent magnet synchronous motor(PMSM),a fast predictivecurrent control(FPCC)requiring only one prediction and comparison is obtained.After furthersimplification on FPCC,this paper proposes a model-free PCC(MFPCC),which only requires singlecurrent sampling during one control period.This method has low computational burden,strongrobustness,good steady state performance and quick response.A comparative study between FPCC andMFPCC is carried out in terms of parameter sensitivity,steady state performance and dynamic response.The simulation and experimental results confirm the effectiveness of the proposed methods.
作者
张虎
张永昌
刘家利
高素雨
Zhang Hu;Zhang Yongchang;Liu Jiali;Gao Suyu(Power Electronics and Motor Drive Engineering Research Center of Beijing North China University of Technology Beijing 100144 China)
出处
《电工技术学报》
EI
CSCD
北大核心
2017年第2期180-187,共8页
Transactions of China Electrotechnical Society
基金
国家自然科学基金(51577003
51207003)
北京市自然科学基金(3162012)
北京市科技新星计划(xx2013001)资助项目
关键词
永磁同步电机
快速预测电流控制
无模型预测电流控制
参数敏感性
稳态和动态性能
Permanent magnet synchronous motor (PMSM)
fast predictive current control
model-free predictive current control
parameter sensitivity
steady state and dynamic performance
