劳文洁, 史林军, 王伟, 杨冬梅, 吴峰, 林克曼. 基于改进DDPG的变速抽蓄机组参与系统调频研究[J]. 太阳能学报, 2024, 45(3): 240-250. DOI: 10.19912/j.0254-0096.tynxb.2022-1846
引用本文: 劳文洁, 史林军, 王伟, 杨冬梅, 吴峰, 林克曼. 基于改进DDPG的变速抽蓄机组参与系统调频研究[J]. 太阳能学报, 2024, 45(3): 240-250. DOI: 10.19912/j.0254-0096.tynxb.2022-1846
Lao Wenjie, Shi Linjun, Wang Wei, Yang Dongmei, Wu Feng, Lin Keman. RESEARCH ON FREQUENCY CONTROL OF POWER SYSTEM WITH DFIM-PSH BASED ON IMPROVED DDPG[J]. Acta Energiae Solaris Sinica, 2024, 45(3): 240-250. DOI: 10.19912/j.0254-0096.tynxb.2022-1846
Citation: Lao Wenjie, Shi Linjun, Wang Wei, Yang Dongmei, Wu Feng, Lin Keman. RESEARCH ON FREQUENCY CONTROL OF POWER SYSTEM WITH DFIM-PSH BASED ON IMPROVED DDPG[J]. Acta Energiae Solaris Sinica, 2024, 45(3): 240-250. DOI: 10.19912/j.0254-0096.tynxb.2022-1846

基于改进DDPG的变速抽蓄机组参与系统调频研究

RESEARCH ON FREQUENCY CONTROL OF POWER SYSTEM WITH DFIM-PSH BASED ON IMPROVED DDPG

  • 摘要: 在挖掘双馈型抽水蓄能(DFIM-PSH)机组调频能力的基础上,提出一种基于改进深度确定性策略梯度(DDPG)算法的系统频率控制方法。首先,基于所确定的DFIM-PSH机组在发电与抽水工况下的频率控制环节,构建考虑风电接入的含DFIMPSH单区域系统频率控制模型。其次,在考虑机组运行约束的基础上以最小化系统频率偏差及调频出力为目标,引入DDPG算法对各机组的AGC控制指令进行优化。通过在预学习中同时引入随机外部扰动与模型参数变化,提高AGC控制器在具有强不确定性环境中的适应性。最后,在仿真验证DFIM-PSH调频优势的基础上,在不同风电接入及扰动等多场景进行仿真分析,结果表明,所提频率控制方法能有效改善新型电力系统的频率特性且具有强鲁棒性。

     

    Abstract: On the basis of exploring the frequency regulation ability of doubly-fed induction machine pumped storage hydro(DFIMPSH)unit,a system frequency control method based on improved deep deterministic policy gradient(DDPG)algorithm is proposed.Firstly,based on the frequency regulation links of DFIM-PSH in generating and pumping modes,the frequency control model of singleregion system with DFIM-PSH considering wind power integration is constructed. Secondly,aiming at minimizing the system frequency deviation and unit’s output for frequency regulation,improved DDPG algorithm is introduced to optimize the AGC instructions of each unit taking the unit’s operating constraints into account. By introducing random external disturbances and model parameter changes in the pre-learning process,the adaptability of AGC controller in environment with strong uncertainty is improved. Finally,on the basis of verifying the advantages of DFIM-PSH,simulations under different wind power integration and disturbance scenarios are carried out.The results show that the proposed frequency control method can effectively improve the frequency characteristics of power system and has strong robustness.

     

/

返回文章
返回