研究目的
To propose an innovative methodology for modeling intermittent PV generation and smart meter readings to reduce complexities in power system reliability assessment.
研究成果
The proposed innovative methodology for PV and load modeling achieves high accuracy (over 99.5% for generation and 97.5% for load) in reliability assessment, reducing complexity and computational effort. It is robust for applications in power system security and risk modeling, though load linearization has minor inaccuracies. Future work could improve load modeling accuracy and extend to other renewable sources.
研究不足
The linearization process for load profiles has a slightly lower accuracy (97.5%) compared to PV profiles (99.5%), due to additive errors from consumer clustering. The study is based on specific datasets (UK weather and Irish smart meter data) and may not generalize to other regions. Computational complexity is reduced but not eliminated.
1：Experimental Design and Method Selection:
The methodology involves simulating PV generation profiles using weather data and linearizing both PV and load profiles to reduce complexity. A stochastic model based on historical weather data is used for PV simulation, and an extended k-means clustering algorithm is applied to smart meter data for load profiling. Linearization is performed using Taylor series expansion and optimized with particle swarm optimization to minimize energy variation.
2：Sample Selection and Data Sources:
Hourly weather data (air temperature and total solar radiation) for Belfast city is simulated using the UKCP09 weather generator. Load demand data is sourced from the 'ISSDA CER' Smart metering dataset, containing half-hourly energy consumption records of nearly 6000 consumers.
3：List of Experimental Equipment and Materials:
No specific equipment or materials are mentioned; the study relies on computational models and datasets.
4：Experimental Procedures and Operational Workflow:
PV generation profiles are simulated from weather data, interpolated to half-hourly resolution, and linearized by selecting threshold points, applying Taylor series linearization, interpolating intermediate points, and optimizing with PSO. Load profiles are clustered and linearized similarly. The linearized profiles are used in reliability assessments via non-sequential Monte Carlo simulation and optimal power flow analysis.
5：Data Analysis Methods:
Reliability is assessed using the Expected Energy Not Supplied (EENS) index, calculated from load curtailment and system state probabilities. Accuracy is validated by comparing linearized profiles with original profiles.
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
