研究目的
Aiming at the correlation of light intensity and load fluctuations, this paper proposes a new photovoltaic-load model method for building correlations.
研究成果
In this paper, we focus on the correlation between light intensity and load fluctuations, and propose a photovoltaic-load model method that involves correlation. Proposed the non-parametric kernel density estimation method to analyze the probability distribution of light intensity and node load, without estimating the distribution type, and able to simulate the distribution of variable historical data more accurately; A photovoltaic-load correlation model was established by combining the Copula function with the Latin hypercube sampling method, and the correlation between the light intensity and the nodal load was taken into account.
研究不足
Not explicitly mentioned in the provided text.
1：Experimental Design and Method Selection:
The non-parametric kernel density estimation method was used to analyze the probability distribution of light intensity and node load, and the corresponding bandwidth was obtained. The correlation between light intensity and node load was analyzed by using Kendall rank correlation coefficient. The load nodes were divided into strong light intensity related nodes and weakly correlated nodes. For the node with strong light intensity, the Copula function is used to establish the photovoltaic-load correlation model. For the weakly correlated nodes of light intensity, the Latin hypercube sampling method was used to establish the load fluctuation model.
2：Sample Selection and Data Sources:
Historical data of light intensity and node load in a city's power grid.
3：List of Experimental Equipment and Materials:
Not explicitly mentioned.
4：Experimental Procedures and Operational Workflow:
Input historical data of light intensity and node load, use non-parametric probability kernel density estimation method to calculate bandwidth estimation value; calculate the Kendall rank correlation coefficient between node load variables and light intensity, and divide load nodes into two categories; for the light-intensity and light-intensity correlated load nodes, establish a photovoltaic-load correlation model using the Copula function; for the weakly correlated light load nodes, establish a random load model using the Latin hypercube sampling method.
5：Data Analysis Methods:
The non-parametric kernel density estimation method and Kendall rank correlation coefficient were used for data analysis.
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
