氢燃料汽车双向DC-DC变换器改进模型预测控制

Model predictive control for bidirectional DC-DC converter of hydrogen fuel vehicles

双向交错并联DC-DC(BDC)变换器是氢燃料电池汽车实现供电可靠性和能量回收的重要设备,采用传统控制方法时BDC变换器面临着响应速度慢、稳定性不高、输出电流纹波大等问题,针对上述问题,本文采用一种带约束条件改进模型预测电流控制方法。该策略针对BDC变换器两种不同工作模式分别建立其数学模型,基于矢量工作原理搭建BDC变换器不同工作模式改进电流预测模型;然后针对模型预测控制过程开关抖动频繁的问题,对成本函数进行优化设计,约束条件中加入控制变量增量;为解决输出电流纹波问题,改进模型预测控制策略通过在线计算开关占空比,得出矢量作用时间,设计成本函数实现控制目标;实验和仿真对比结果显示传统电流控制方法响应时间和电流纹波分别为0.1 s和5 A,改进MPC模型预测控制响应时间和电流纹波分别为0.02 s和1.5 A,实验和仿真对比结果表明带约束条件模型预测电流控制具有更好的动态响应和稳定性能,验证该算法的有效性。

The bidirectional interleaved parallel DC-DC(BDC)converter is an important device for hydrogen fuel cell vehicle power supply reliability and energy recovery.The BDC converter experiences slow response speed,low stability,and output current ripple when using traditional control methods.Aiming at the above problems,a constrained model predictive current control is proposed.First,mathematical models for the various working modes of the BDC converter are established,and an improved current prediction model for the various working modes of the BDC converter based on the vector working principle is constructed.The cost function is then optimized to solve the problem of frequent switch jitter in the model predictive control process,and the control variable increment is added to the constraint condition;to solve the output current ripple problem,the model predictive control strategy is improved by calculating the switch duty cycle online.Finally,calculate the vector action time and create the cost function to achieve the control objective.The traditional current control method's response time and current ripples are 0.1 s and 5 A,respectively.In contrast,the improved MPC model predictive control response time and current ripples are 0.02 s and 1.5 A.The experimental and simulation comparison results show that Model predictive current control with constraints has a better dynamic response and stable performance,which verifies the algorithm's effectiveness.