首页|基于向量投影测度的区间型组合预测模型

基于向量投影测度的区间型组合预测模型

扫码查看
区间型组合预测模型的研究核心是确定各单项预测方法的权重.鉴于向量投影是一个综合性测度,既能反映两个向量指标对象之间距离,又能反映他们之间的夹角,本研究提出了一种新的向量标准化投影测度公式.在向量标准化投影测公式的基础上,定义了区间数向量标准化投影测度的概念,用以描述两个区间数向量的接近程度,以组合预测区间数序列和实际区间数序列的标准化投影测度为最优化准则,构建了两个基于投影测度的全新区间型组合预测模型.并通过两个实例进行了分析:(1)与已有文献的组合预测方法的结果作对比研究,验证了本法所构建的组合预测模型的有效性;(2)对我国 2010-2022 风力发电量序列进行拟合和预测建模,结果表明本法构建的组合预测模型有更好的拟合和预测效果,具有一定的可行性和实用性.
INTERVAL COMBINATION FORECASTING MODEL BASED ON GENERALIZED PROJECTION MEASURE
The core of interval combined forecasting model is to determine the weight of each single forecasting method.Since vector projection is a comprehensive measure,which can reflect both the distance and angle between two vector index objects,this paper proposes a new vector standardized projection measure formula.On the basis of vector standardized projection measure formula,the concept of interval number vector standardized projection measure is defined to describe the proximity of two interval number vectors.Two new interval type combination prediction models based on projection measure are constructed by taking the standardized projection measure of combination prediction interval number sequence and actual interval number sequence as the optimization criteria.Two examples are analyzed:(1)The effectiveness of the combined forecasting model built in this paper is verified by comparing the results of the combined forecasting methods with those of the existing literature;(2)The fitting and forecasting model of China's 2010-2022 wind power generation series shows that the combined forecasting model built in this paper has better fitting and forecasting effect,and has certain feasibility and practicability.

interval numberstandardized projectioncombination modelwind power generationforecast

马虹、谢小军、薛申芳、龙彩燕

展开 >

广东金融学院金融数学与统计学院,广东,广州 510521

广州工商学院通识教育学院,广东,广州 510850

区间数 标准化投影 组合模型 风力发电量 预测

国家自然科学基金项目广东省2022年度普通高校重点科研平台和项目广州工商学院2021年校级科研项目

615720162022ZDZX1037KA202132

2024

井冈山大学学报(自然科学版)
井岗山大学

井冈山大学学报(自然科学版)

影响因子:0.298
ISSN:1674-8085
年,卷(期):2024.45(1)
  • 18