研究目的
To address the computational burden of Quasi-static time-series (QSTS) simulations for distribution system feeders of any complexity by expanding the vector quantization approach, thereby maintaining high computational time reduction.
研究成果
The adapted vector quantization algorithm can model realistic size, complex feeders with multiple profiles and controllable elements, providing a significant computational time reduction (98.7%) for QSTS simulations. This advancement eliminates the main limitation to the wide use of QSTS simulation by the industry, offering great insights for system planners as DER penetration increases.
研究不足
The algorithm's effectiveness is dependent on the sparsity of the space due to the correlation between controllable elements and the precision of the profiles. The indexing matrix can become extremely large, requiring significant memory.
1：Experimental Design and Method Selection:
The vector quantization algorithm groups similar power flow solutions in clusters to avoid the iterative power flow solver, storing solutions for consequent time steps.
2：Sample Selection and Data Sources:
A realistic distribution feeder with 2969 buses, 3 load/PV profiles, and 8 voltage regulating elements was modeled in OpenDSS through a MATLAB interface.
3：List of Experimental Equipment and Materials:
A Window 10 computer with 32GB of memory and a
4：50GHz processor was used for simulation. Experimental Procedures and Operational Workflow:
The algorithm cycles through determining the power flow solution and whether an action is taken by controllable elements, using quantization logic to bypass the iterative AC power flow solver.
5：Data Analysis Methods:
The computational time and accuracy of the simulation were analyzed for different vector quantization cases.
独家科研数据包，助您复现前沿成果，加速创新突破
获取完整内容
