车载Buck-LLC变换器双闭环控制策略研究

Research on Dual-Closed-Loop Control Strategy for On-Board Buck-LLC Converter

摘要: 针对车载辅助电源在宽输入电压范围下对高动态响应和高稳定性的需求，本文以Buck-LLC两级式变换器为研究对象，提出一种基于模型预测控制与自抗扰控制的双闭环复合控制策略。首先，利用后级LLC谐振电路在准谐振状态下的直流变压器特性，提出一种简化模型预测控制内环算法，通过输出电压折算值替代中间母线电压采样，实现了电感电流的无差拍跟踪并减少了采样电路。其次，针对外环ADRC控制器，引入粒子群算法进行控制参数的离线全局寻优，解决了传统自抗扰控制参数整定困难的问题。Matlab/ Simulink仿真结果表明：相比于传统双PI控制，该策略在输入电压剧烈波动及负载突变工况下，超调量显著减小，调节速度变快。此外，该策略对电感参数摄动(±20%)表现出极强的容错能力，有效克服了传统MPC对模型精度依赖性高的缺陷，具有较高的工程应用价值。

Abstract: To address the requirements for high dynamic response and stability of on-board auxiliary power supplies across a wide input voltage range, this paper investigates the Buck-LLC two-stage converter and proposes a dual-closed-loop composite control strategy based on Model Predictive Control (MPC) and Active Disturbance Rejection Control (ADRC). Firstly, leveraging the “DC transformer” characteristics of the subsequent LLC resonant circuit in the quasi-resonant state, a simplified MPC inner-loop algorithm is proposed. By utilizing the converted output voltage value instead of intermediate bus voltage sampling, deadbeat tracking of the inductor current is achieved while the sampling circuitry is simplified. Secondly, for the ADRC outer-loop controller, the Particle Swarm Optimization (PSO) algorithm is introduced for offline global parameter optimization, overcoming the tuning difficulties associated with traditional ADRC. Matlab/Simulink simulation results indicate that, compared with traditional dual-PI control, the proposed strategy significantly reduces overshoot and shortens regulation time under drastic input voltage fluctuations and load transients. Furthermore, the strategy exhibits excellent robustness against inductor parameter perturbations (±20%), effectively addressing the high model-accuracy dependency of traditional MPC and demonstrating significant engineering application value.