摘要
为了降低双向电子变压器输出端的DC-DC变换电路中的开关损耗,目前广泛采用的是电容-电感-电感-电容(CLLC)谐振电路。其中谐振参数的选取直接影响电子变压器的输出性能。为此,引入量子粒子群优化(QPSO)算法对谐振参数进行智能优化设计。该算法能够对存在不可导点的目标函数进行优化,通过引入量波动方程,增加了粒子搜索能力,解决了粒子群优化(PSO)算法局部收敛等问题。其优化过程首先利用谐振点分区法对变换器的参数进行建模,然后分析和研究谐振电感、励磁电感与谐振网络直流增益、输入阻抗角以及相对损耗率的直接关系,得到参数优化的约束条件。在此基础上,利用QPSO算法得到电路相对损耗最小的谐振网络参数。最后,以CLLC谐振变换电路为例,通过实验验证了在宽输入电压下、全负载范围内,该优化设计方法与传统基波分析法与时域法相比,效率均提高了1%左右,样机最高效率达97%,从而证明了所提优化方法的有效性。
In order to reduce the switching loss of DC-DC converter at the output port of bi-directional electronic transformer, a capacitor-inductor-inductor-capacitor(CLLC)resonant circuit is widely used. One of the problems is the selection of resonance parameters, which will directly affect the output performance of electronic transformer. A quantum particle swarm optimization(QPSO) algorithm is introduced to implement the intelligent optimization design of resonance parameters. The algorithm can optimize the objective function with non-differentiable points, and the quantity wave equation is used to enhance the particle search ability and to solve the local convergence problem of particle swarm optimization(PSO) algorithm. In the optimization process,a resonance point partition method is applied to establish the mathematical model of DC-DC converter, and the relationship between resonance parameters and the characteristics of the resonance network including the DC gain, input impedance angle and relative loss rate are analyzed to obtain the constraints of parameter optimization. On this basis, QPSO is used to obtain the resonance network parameters of the relative minimum loss of the circuit. Finally, taking CLLC resonance converter as an example, it is verified by experiments that in the wide-input-voltage and full-load-range conditions, the efficiency of the optimized design method based on QPSO algorithm is about 1% higher than that based on the traditional fundamental analysis(FHA) method and time domain method, and the highest efficiency of prototype reaches 97%. Thus the effectiveness of the optimization method proposed is proven.
作者
江友华
帅禄玮
吴琦娜
曹以龙
JIANG Youhua;SHUAI Luwei;WU Qina;CAO Yilong(College of Electronics and Information Engineering,Shanghai University of Electric Power,Shanghai 200090,China;Shinan Electricity Supply Company of State Grid Shanghai Municipal Electric Power Company,Shanghai 200233,China)
出处
《电力系统自动化》
EI
CSCD
北大核心
2020年第11期171-180,共10页
Automation of Electric Power Systems
基金
国家自然科学基金资助项目(61202369)
上海市科技创新行动计划资助项目(19DZ1205402)。
关键词
双向电子变压器
电容-电感-电感-电容(CLLC)变换器
谐振参数优化
谐振点分区
量子粒子群优化
bi-directional electronic transformer
capacitor-inductor-inductor-capacitor(CLLC)converter
resonance parameter optimization
division of resonance points
quantum particle swarm optimization
