Modeling and Stability Analysis for Grid-Forming Converter in a Renewable Energy Generation System

Yankai XING,Guangdou Zhang,Baolu Wang,Jian Li,Olusola Bamisile,Qi Huang
DOI: https://doi.org/10.2139/ssrn.4356272
2023-01-01
Abstract:The fast response and low inertia characteristics of converter-based generation (CBG) lead to a new stability issue that limits the development of renewable energy. The experimental tests validated in this paper confirm that although the grid-forming type converter (GFM) can improve the stability of CBG, the wideband oscillation problems are still risks within different Short-Circuit Ratio (SCR) conditions. This is due to the fact is that, there is no unified and effective way to setting the parameters in existence control strategy. The unsuitable selection of the output and input variable in linearized model of GFM converter leads to a failure of coefficients tuning which brings the instability to the system. In order to optimization parameters tuning process, a benchmark model is adopted in this paper, and the linearization model is updated by selecting suitable variables and detailed with considering the influence of resistance of the converter, etc. Based on this model, the stability margin related to the parameters of system is analyzed. Furthermore, based on this model, in order to consider more dynamics and make the converter controller focus on the topologies and scenarios which lack stabilitythemselves, the different grid scenarios are emulated to design a controller using Deep Reinforcement Learning (DRL) from offline to online to capture more dynamics. That is, the cascaded voltage and current control is tuned by Deep Deterministic Policy Gradient (DDPG) algorithm. Furthermore, the novel grid-forming control structure is proposed based on same algorithm. The validation tests indicate that the proposed strategy provides more essential for the power system stability in various grid conditions.
What problem does this paper attempt to address?