论文部分内容阅读
针对直流微电网系统的并网仿真问题,提出了一种基于模糊径向基(fuzzy radial basis function)神经网络的直流微电网动态等效建模方法。利用直流微电网并网接入母线端量测电压、电流和功率数据,构建基于径向基神经网络的并网等效模型。利用模糊聚类规则训练人工神经网络,并利用改进细菌觅食算法进行神经网络结构和参数的辨识。提出的模型能够反映直流微电网的整体动态特性,能够跟踪微电网不同运行方式下的网络拓扑和结构变化。该等效模型及建模方法为直流微电网及其并网仿真分析提供了一种思路和解决办法,仿真结果验证了所提方法的合理性和可靠性。
Aiming at the grid-connected simulation problem of DC microgrid system, a dynamic equivalent modeling method based on fuzzy radial basis function neural network is proposed. The voltage, current and power data of DC microgrid are used to measure the grid-connected equivalent model based on RBF neural network. Using fuzzy clustering rules to train artificial neural network and using improved bacteria foraging algorithm for neural network structure and parameter identification. The proposed model can reflect the overall dynamic characteristics of the DC microgrid and can track the network topology and structure changes under different operating modes of the microgrid. The equivalent model and modeling method provide a new idea and solution for the DC micro-grid and its grid-connected simulation analysis. The simulation results verify the rationality and reliability of the proposed method.