| 引用本文: | 张梦琪,李永刚,孙庚,吴滨源,刘淇玉,张驰.基于DBN-ELM的构网型并网逆变器控制参数自适应调整方法[J].电力自动化设备,2024,44(4):111-118 |
| ZHANG Mengqi,LI Yonggang,SUN Geng,WU Binyuan,LIU Qiyu,ZHANG Chi.Adaptive adjustment method for control parameters of grid-forming inverter based on DBN-ELM[J].Electric Power Automation Equipment,2024,44(4):111-118 |
|
| 摘要: |
| “双高”电力系统中电网阻抗呈现宽范围时变特性，构网型并网逆变器控制参数缺乏自适应调整能力，存在失稳风险。对此，提出一种基于深度置信网络-极限学习机的构网型并网逆变器控制参数自适应调整方法。建立闭环极点映射模型，利用深层架构对控制参数与关键极点之间的映射关系进行训练；通过训练好的闭环极点映射模型预测得到相应的关键极点，识别出关键极点最接近参考极点时构网型并网逆变器的控制参数；通过自适应调整控制参数，确保系统在电网阻抗变化时跟踪参考极点，实现自适应稳定控制。理论分析和仿真结果均表明，所提方法能够实现控制参数的自适应调整，有效提高构网型并网逆变器对电网阻抗变化的适应性。 |
| 关键词: 构网型并网逆变器 自适应调整 深度置信网络-极限学习机 复矢量建模 电网阻抗 |
| DOI：10.16081/j.epae.202308008 |
| 分类号:TM464 |
| 基金项目:国网辽宁省电力有限公司管理科技项目(2022YF-36) |
|
| Adaptive adjustment method for control parameters of grid-forming inverter based on DBN-ELM |
|
ZHANG Mengqi1, LI Yonggang1, SUN Geng2, WU Binyuan1, LIU Qiyu1, ZHANG Chi1
|
|
1.State Key Laboratory of Alternate Electrical Power System with Renewable Energy Sources, North China Electric Power University, Baoding 071003, China;2.Fuxin Power Supply Subsidiary Company of State Grid Liaoning Electric Power Co.,Ltd.,Fuxin 123000, China
|
| Abstract: |
| In the “double-high” power system, the grid impedance exhibits a wide range of time-varying characteristics. The grid-forming control parameters of grid-forming inverter lack the adaptive adjustment ability, resulting in the risk of instability. To solve this problem, an adaptive adjustment method for control parameters of grid-forming inverters based on deep belief network and extreme learning machine is proposed. A closed-loop pole mapping model is established and the mapping relationship between control parameters and key poles is trained by using the deep architecture. Then, the corresponding key poles are predicted by the trained closed-loop pole mapping model, and the control parameters of the grid-forming inverter are identified when the key poles are closest to the reference poles. By adjusting the control parameters adaptively, the system can keep track the reference poles when the grid impedance changes, and realize the adaptive stability control. It is shown by both theoretical analysis and simulative results that the proposed method can realize the adaptive adjustment of control parameters, and effectively improve the adaptability of the grid-forming inverter to the changes of the grid impedance. |
| Key words: grid-forming inverter adaptive adjustment deep belief network-extreme learning machine complex vector modeling grid impedance |
