电工技术学报2024，Vol.39Issue(10)：3093-3104,3140,13.DOI:10.19595/j.cnki.1000-6753.tces.230361
双有源全桥变换器无电流传感器鲁棒预测控制
An Enhanced Current Sensorless Predictive Control for Dual Active Bridge Converter
摘要
Abstract
With the development of renewable energy power generation,electrified transportation,and energy storage systems,the power system is entering a new stage of coordinated and optimized operation of source-grid-load-storage.The energy router based on power electronic devices can flexibly access and efficiently use various renewable energy sources,which has become the core equipment of the new power grid.The dual active full-bridge(DAB)DC-DC converter has the advantages of high power density,bidirectional energy transmission,natural electrical isolation,and easy realization of soft switching.Therefore,it has gradually become one of the important topologies on the DC side of energy routers. In the new power grid,the increased integration of renewable energy sources presents challenges of complex operating conditions in power systems and frequent switching of various energy storage equipment.Therefore,energy routers must have good dynamic performance.The traditional PI control employed in DAB converter systems tends to prolong dynamic performance due to the limitation of the equivalent low-pass filtering characteristics of the controller,which has become a bottleneck for the dynamic improvement of the DAB converter.This paper proposes a whole process model predictive control strategy.Firstly,the prediction model of the DAB converter is established.For the start-up process of the DAB converter,the dynamic performance optimization problem,formulated as an equality constraint optimization problem,is solved by KKT conditions.For the transient process of the DAB converter,the optimal phase shift ratio is determined based on the mathematical analysis of the cost function. The above whole process model predictive control strategy can effectively improve the dynamic performance of the DAB converter system.However,when applying the MPC to the DAB converter system,the high requirements on model parameters and sampling accuracy need to be considered.Therefore,a current sensorless robust control strategy based on the Kalman filter algorithm is proposed.The DAB converter system's discrete state space model and state estimation model are established.The system state variables are reconstructed,and the observation coefficients are improved to linearize the system.The optimal estimation value is obtained according to the Kalman filter principle,and the sensorless robust predictive control is realized. In summary,this paper proposes a whole process model predictive control strategy for the DAB converter system,providing a detailed analysis of model parameter sensitivity and sampling accuracy under the MPC strategy.A current sensorless robust predictive control strategy based on the Kalman filter is proposed.The theoretical analysis and experimental verification of the proposed method reveal the following advantages. (1)Compared with the classical two-stage start-up strategy and PI control strategy,the proposed full-process model predictive control strategy greatly improves the dynamic performance of the DAB converter system during start-up speed and transient processes. (2)The proposed current-sensorless robust predictive control strategy can maintain good dynamic performance without a current sensor,overcoming the parameter dependence of the MPC strategy.The proposed control strategy is robust to model parameter error and sampling noise.关键词
双有源全桥/DC-DC/变换器/模型预测控制/动态响应提升/无电流传感器控制鲁棒控制Key words
Dual active bridge DC-DC converter/model predictive control/dynamic response enhancement/current sensorless control/robust control分类
信息技术与安全科学引用本文复制引用
孙孝哲,张祯滨,韩明昊,李真,刘晓哲..双有源全桥变换器无电流传感器鲁棒预测控制[J].电工技术学报,2024,39(10):3093-3104,3140,13.基金项目
深圳市科技计划(JCYJ20210324132616040)、广东省基础与应用基础研究基金(2022A1515240028)和国家自然科学基金(51977124,52277192,52277191)资助项目. （JCYJ20210324132616040）
