Safety Assessment Method for Wind-Solar Systems Based on Improved Latin Hypercube Sampling

In response to operational stability challenges stemming from the inherent randomness and volatility of high-penetration wind-solar power sources in the grid, it is crucial to statistically quantify system uncertainties and evaluate network risks. This paper proposes a safety assessment method for wind-solar systems using an enhanced Latin hypercube sampling approach. First, a renewable energy output sample space is generated via slice sampling (SS) based on a non-parametric kernel density estimation (KDE) model. Subsequently, this sample space incorporates system random variables (such as load demand) alongside uncertainty samples from renewable energy, utilizing Latin hypercube sampling (LHS) to derive the final sample set. Risk assessment of the wind-solar system is then performed using the developed safety assessment indicators. Simulation results demonstrate that the proposed combined sampling method significantly reduces sample size while satisfying accuracy requirements, thereby improving computational efficiency compared to alternative algorithms.