胡治国, 胡宏伟, 艾永乐,等.LLC谐振变换器变论域模糊神经PI控制研究[J].河南理工大学学报（自然科学版）,2020,39(5):106-113.

HU Z G, HU H W, AI Y L, et al.Variable-domain fuzzy neural PI control for LLC resonant converter[J].Journal of Henan Polytechnic University(Natural Science) ,2020,39(5):106-113.

LLC谐振变换器变论域模糊神经PI控制研究

胡治国1, 胡宏伟1, 艾永乐1, 褚艳伟2

1.河南理工大学 电气工程与自动化学院，河南 焦作 454000;2.珠海格力电器股份有限公司，广东 珠海 519000

摘要:LLC谐振变换器具有高功率密度、高效率，且易实现全负载范围内软开关等优点，已获得广泛应用，但由于LLC谐振变换器具有强非线性和难以获得精确数学模型的特点，致使传统PI控制无法实现良好的控制性能。为改善LLC谐振变换器控制性能，本文在分析模糊控制、神经网络控制和变论域原理基础上，设计一种变论域模糊神经PI控制器。在模糊控制的基础上，通过引入神经网络，对模糊隶属度函数的分布进行优化，同时通过引入变论域伸缩因子提高模糊规则的利用率。最后对设计的变论域模糊神经PI控制器在LLC谐振变换器闭环控制中的性能进行研究，结果表明，本文设计的变论域模糊神经PI控制具有超调小、动态响应快等优点，其控制性能优于传统PI控制和模糊PI控制。

关键词:LLC谐振变换器;模糊控制;人工神经网络;变论域原理;伸缩因子

doi:10.16186/j.cnki.1673-9787.2020.5.16

基金项目:河南省科技攻关项目（172102310569）

收稿日期:2019/12/08

修回日期:2020/03/19

出版日期:2020/09/15

Variable-domain fuzzy neural PI control for LLC resonant converter

HU Zhiguo1, HU Hongwei1, AI Yongle1, CHU Yanwei2

1.School of Electrical Engineering and Automation， Henan Polytechnic University， Jiaozuo  454000 ， Henan， China;2.Gree Electric Appliances Inc. of Zhuhai， Zhuhai  519000 ， Guangdong，China

Abstract:LLC resonant converter has many advantages such as high power density， high efficiency， easy to realize soft switch in full load range， etc. ， and has been widely used. However， due to the characteristics of LLC resonant converter such as strong nonlinearity and difficulty in obtaining accurate mathematical model， the traditional PI control cannot get good control performance. To improve the control performance of LLC resonant converter， a fuzzy neural PI controller was designed based on the theory of fuzzy control， neural network control and variable domain. On the basis of fuzzy control， the distribution of fuzzy membership function was optimized by introducing neural network， and the utilization rate of fuzzy rules was improved by introducing scaling factor of variable domain theory. Finally， the performance of the designed variable -domain fuzzy neural PI controller in the closed-loop control of LLC resonant converter was studied. The results showed that the variable-domain fuzzy neural PI controller designed in this paper had the advantages of small overshoot and fast dynamic response ，and its control performance was better than that of conventional PI control and fuzzy PI control.

Key words:LLC resonant converter;fuzzy control;artificial neural network;variable domain theory;scaling factor

LLC谐振变换器变论域模糊神经PI控制研究_胡治国.pdf