| 引用本文: | 车玉龙,王晓茹,吕晓琴,康永强.基于最优箱宽直方图的牵引变电所负荷概率建模方法[J].电力自动化设备,2020,40(12): |
| CHE Yulong,WANG Xiaoru,Lü Xiaoqin,KANG Yongqiang.Load probability modeling method of traction substation based on optimal bin width histogram[J].Electric Power Automation Equipment,2020,40(12): |
|
| 摘要: |
| 电气化铁路牵引负荷受众多随机因素的影响，较难用已知的参数估计模型来统一描述其概率模型。直方图是最简单直观且应用最广泛的一种非参数估计方法，但由其得到的概率密度形状极易受箱宽(或箱数)的影响。为得到牵引负荷概率分布的最佳直方图，提出一种基于最优箱宽建立其概率模型的直方图方法。由Poisson点事件构造直方图，通过平均积分平方误差定义价值函数，利用Sturges规则和平均移位直方图思想计算最优箱宽。采用3种不同采样间隔的牵引变电所负荷测试数据，通过概率分布形状和后验检验，与4种传统计算直方图箱宽的经验规则方法进行比较，验证了所提方法的有效性、适应性和稳健性。 |
| 关键词: 牵引负荷 概率模型 直方图 价值函数 最优箱宽 平均移位 |
| DOI：10.16081/j.epae.202012006 |
| 分类号:TM922.3 |
| 基金项目:甘肃省高等学校创新能力提升项目(2019B-049)；甘肃省高等学校创新基金资助项目(2020A-036)；兰州交通大学青年科学研究基金资助项目(20200013) |
|
| Load probability modeling method of traction substation based on optimal bin width histogram |
|
CHE Yulong1,2, WANG Xiaoru2, Lü Xiaoqin2, KANG Yongqiang3
|
|
1.School of Automation and Electrical Engineering, Lanzhou Jiaotong University, Lanzhou 730070, China;2.School of Electrical Engineering, Southwest Jiaotong University, Chengdu 611756, China;3.School of New Energy and Power Engineering, Lanzhou Jiaotong University, Lanzhou 730070, China
|
| Abstract: |
| As the traction load in electric railway is affected by many random factors, it is difficult to uniformly describe the load probability model by the known parameter estimation models. Histogram is the simplest, most intuitive and widely used non-parametric estimation method, but the shape of the probability density obtained from histogram is extremely susceptible to bin width(or bin number). In order to obtain the optimal histogram of traction load probability density, a histogram method based on the optimal bin width is proposed to build its probability model. The histogram is constructed by Poisson point events, the value function is defined by mean integrated squared error, and the optimal bin width is calculated by Sturges rule and the average shift histogram idea. Based on the load test data of traction substations with three different sampling intervals, through the shape of probability distribution and the posterior test, the vali-dity, adaptability and robustness of the proposed method are verified by comparing with four traditional empirical rule methods for bin width calculation. |
| Key words: traction load probability model histogram value function optimal bin width average shift |
