基于非线性扩张状态观测器的交错并联Buck变换器无模型预测控制

Model-Free Predictive Control for Three-Phase Interleaved Buck Converter Base Nonlinear Extended State Observer

针对交错并联Buck变换器在复杂环境下受到内外部扰动导致控制性能下降问题,提出一种基于非线性扩张状态观测器的无模型预测控制策略(NESO-MFPC)。建立超局部模型来替换原有先验物理模型,降低控制系统模型精度依赖。设计非线性扩张状态观测器估计超局部模型中未知干扰部分,并补偿到基于超局部模型的预测模型中去,提高抗干扰能力。设计无差拍预测控制策略,保证系统的快速动态响应。仿真结果表明,相比于传统控制方法,该控制策略具有更优的动稳态性能和鲁棒性。

To solve the problem of control performance degradation caused by internal and external disturbances of three-phase interleaved buck converter in working conditions,a model-free predictive control strategy based on a nonlinear extended state observer(NESO-MFPC)was proposed.A ultra-local model was established to replace the original physical model to reduce the dependency of system model accuracy.A nonlinear extended state observer was designed to estimate the uncertain part of the ultra-local model,which improved anti-disturbance ability by compensating disturbance to prediction model.Model-free predictive control was adopted based on the deadbeat control method to achieve better dynamic performance.The simulation results validate the better dynamic and steady-state performance and stronger robustness of NESO-MFPC compared with conventional method.