基于改进非参数核密度估计和拉丁超立方抽样的电动公共客车负荷模型

Electric Public Bus Load Model Based on Improved Kernel Density Estimation and Latin Hypercube Sampling

为了建立更精确的电动公共客车充电负荷模型,对起始荷电状态和起始充电时间的概率分布和抽样方法进行研究。首先针对概率分布,提出采用边界核的自适应非参数核密度估计算法,该方法无需概率分布的任何假设,能有效解决边界偏差和缺乏局部适应性的问题,从而提高了起始荷电状态和起始充电时间概率分布的准确度和适应性。其次针对抽样环节,提出结合三次样条插值法的改进拉丁超立方抽样算法,该方法不仅可以解决累积分布函数的反函数较难求解的问题,而且具有准确度高、抽样规模小等优点。最后,以这两种算法为基础,建立电动公共客车充电负荷模型,并将其与实测数据和传统参数估计方法所得结果进行对比。仿真结果证明了所提方法的适应性和有效性。

To build a more precise charging load model of electric public buses, the probabilitydistribution and sampling method of charging state and charging start time are studied. Firstly, an adaptive kernel density estimation with boundary Kernel algorithm is proposed to build probability distribution models. The proposed algorithm does not require any assumptions about the probability distribution, and can also solve the problems of boundary bias and lacking of local adaptability, which improves the precision and adaptability of probability distribution. Secondly, Latin hypercube sampling with cubic spline interpolation algorithm is proposed to figure out the inverse cumulative distribution function of probability distribution. The proposed algorithm has high precision and sampling efficiency. Finally, based on these two algorithms, electric public bus charging load model is set up. The simulation result demonstrates the effectiveness and adaptability of the proposed method.