WO2020093701A1 - 一种基于AdaBoost-SO的VANETs车辆事故风险预测模型 - Google Patents

一种基于AdaBoost-SO的VANETs车辆事故风险预测模型 Download PDF

Info

Publication number
WO2020093701A1
WO2020093701A1 PCT/CN2019/092462 CN2019092462W WO2020093701A1 WO 2020093701 A1 WO2020093701 A1 WO 2020093701A1 CN 2019092462 W CN2019092462 W CN 2019092462W WO 2020093701 A1 WO2020093701 A1 WO 2020093701A1
Authority
WO
WIPO (PCT)
Prior art keywords
samples
adaboost
sample
data set
vanets
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Ceased
Application number
PCT/CN2019/092462
Other languages
English (en)
French (fr)
Inventor
赵海涛
丁仪
蔡舒祺
张晖
段佳秀
朱洪波
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Nanjing University of Posts and Telecommunications
Original Assignee
Nanjing University of Posts and Telecommunications
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Nanjing University of Posts and Telecommunications filed Critical Nanjing University of Posts and Telecommunications
Publication of WO2020093701A1 publication Critical patent/WO2020093701A1/zh
Anticipated expiration legal-status Critical
Ceased legal-status Critical Current

Links

Images

Classifications

    • GPHYSICS
    • G06COMPUTING OR CALCULATING; COUNTING
    • G06QINFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00Administration; Management
    • G06Q10/04Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
    • GPHYSICS
    • G06COMPUTING OR CALCULATING; COUNTING
    • G06QINFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00Administration; Management
    • G06Q10/06Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
    • G06Q10/063Operations research, analysis or management
    • G06Q10/0635Risk analysis of enterprise or organisation activities

Definitions

  • the invention relates to the field of Internet of Vehicles technology, in particular to a VANETs vehicle accident risk prediction model based on AdaBoost-SO.
  • VANETs vehicle-mounted Ad Hoc networks
  • ITS intelligent transportation systems
  • the traffic accident hotspot prediction Based on the logistic regression method” uses statistical and logistic regression analysis of typical factors to study the relationship between traffic accidents, road types, vehicle types, driver status, weather, and dates. Establish an accident hot spot prediction model.
  • the document “Traffic Big Data Analysis Supporting Network Access Recommendation” developed an intelligent network recommendation system supported by traffic big data analysis. It is recommended that vehicles use the analysis framework to access the appropriate network and allow individual vehicles to automatically access based on the access recommender The internet.
  • the main purpose of the present invention is to solve the problems in the prior art.
  • the present invention provides a VANETs vehicle accident risk prediction model based on AdaBoost-SO.
  • a VANETs vehicle accident risk prediction model based on AdaBoost-SO The steps of establishing the model include:
  • Step 1 Populate the research data set
  • Step 2 SMOTE algorithm is used to balance the samples in the data set, and the discrete features of each sample are encoded with One-Hot;
  • Synthetic Minority Oversampling Technology (SMOTE) algorithm is used to solve the problem of imbalance in the number of samples in each category in the research data set;
  • Step 3 Use trichotomy Adaboost-SO algorithm to train the research data set to obtain the system model
  • the road safety data is randomly divided into training data and test data, and cross-validation is performed 6 times. This method makes full use of all samples. It requires 6 trainings and 6 tests; then , Use trichotomy AdaBoost algorithm to process research data set;
  • Step 4 Import real-time traffic data sets through VANETs to obtain the output of the prediction model
  • the usual implementation scheme includes filling in the average value of available features, special values, The average of similar samples, and directly ignore samples with missing values.
  • the SMOTE algorithm implementation process is:
  • Step 2-1 For each sample x in the minority category, the Euclidean distance is used as a criterion to calculate the distance from all other samples in the minority category to obtain its k nearest sample;
  • Step 2-2 Set the sampling rate N according to the sample imbalance ratio. For each minority sample x, assume that the selected neighboring sample is k, and randomly select several samples from its k neighboring samples;
  • Step 2-3 For each selected neighbor, use the original sample to construct a new sample according to the following formula;
  • step three the specific implementation steps of the six cross-validation are as follows:
  • Step 3-1-1 Divide the entire research data set S into 6 disjoint subsets of the same size; assuming that the number of training samples is m, each subset will have Training samples, the corresponding subsets are ⁇ S 1 , S 2 , S 3 , S 4 , S 5 , S 6 ⁇ ;
  • Step 3-1-2 use one subset as the test set, and then use the other five subsets as the training set;
  • Step 3-1-3 train the model through the training data, use the test data to verify the accuracy of the model and repeat six times;
  • Step 3-1-4 Calculate the average value of 6 evaluation errors as the true classification accuracy of the model.
  • the trichotomy AdaBoost algorithm is used to process the research data set, and the specific implementation steps are as follows:
  • Step 3-2-2 the weights of the training data are initialized as:
  • is the data to be trained
  • the G m (x) error rate is calculated according to the classification results of the training data
  • w mi represents the weight of the i-th sample in the m-th iteration:
  • the denominator Since the weights are normalized in each step, the denominator does not need to be divided by the sum of the sample weights;
  • Step 3-2-4 trichotomy AdaBoost error rate threshold is set to e m And add the positive term x when When, to ensure a m ⁇ 0; coefficient calculation classifier G m (x) according to the error rate e m:
  • Step 3-2-5 construct a linear combination of basic classifiers to obtain the final classifier:
  • the linear combination f (x) implements the weighted voting of M basic classifiers, the f (x) value determines the category of the instance x, and indicates the confidence of the classification, and combines the trained weak classifier into a strong classifier to obtain the risk of vehicle accidents Forecasting model.
  • the beneficial effect of the present invention is that: the system model with a maximum iteration value of 100 guarantees the maximum accuracy of accident prediction under ordinary road conditions, and the system model with a smaller maximum iteration value under special circumstances can improve timeliness. In the prediction, the maximum performance of the system can be exerted.
  • FIG. 1 is a schematic flowchart of the method of the present invention.
  • Figure 2 shows the architecture of trichotomy Adaboost-SO model.
  • a VANETs vehicle accident risk prediction model based on AdaBoost-SO The steps of establishing the model include:
  • Step 1 Populate the research data set.
  • Step 2 Use the SMOTE algorithm to balance the samples in the data set, and encode the discrete features of each sample with One-Hot.
  • the Synthetic Minority Oversampling Technology (SMOTE) algorithm is used to solve the problem of imbalance in the number of samples in each category in the research data set.
  • the SMOTE algorithm implementation process is:
  • Step 2-1 For each sample x in the minority category, the Euclidean distance is used as a criterion to calculate the distance to all other samples in the minority category to obtain its k nearest sample.
  • Step 2-2 Set the sampling rate N according to the sample imbalance ratio. For each minority sample x, assuming that the selected neighboring sample is k, several samples are randomly selected from its k neighboring samples.
  • Step 2-3 For each selected neighbor, use the original sample to construct a new sample according to the following formula.
  • the One-Hot encoding method uses N-bit status registers to encode N states, each state has a separate register bit, and only one bit is valid at any time.
  • Step 3 Use the trichotomy Adaboost-SO algorithm to train the research data set to obtain the system model.
  • Step 3-1-1 Divide the entire research data set S into 6 disjoint subsets of the same size; assuming that the number of training samples is m, each subset will have Training samples, the corresponding subsets are ⁇ S 1 , S 2 , S 3 , S 4 , S 5 , S 6 ⁇ .
  • Step 3-1-2 use one subset as the test set, and then use the other five subsets as the training set.
  • Step 3-1-3 Train the model through the training data, use the test data to verify the accuracy of the model and repeat six times.
  • Step 3-1-4 Calculate the average value of 6 evaluation errors as the true classification accuracy of the model.
  • Step 3-2-2 the weights of the training data are initialized as:
  • is the data to be trained.
  • G m (x) error rate according to the classification results of the training data, w mi represents the weight of the i-th sample in the m-th iteration:
  • the denominator Since the weights are normalized in each step, the denominator does not need to be divided by the sum of the sample weights.
  • Step 3-2-4 trichotomy AdaBoost error rate threshold is set to e m And add the positive term x when When, to ensure a m ⁇ 0; coefficient calculation classifier G m (x) according to the error rate e m:
  • the weights of misclassified samples of the basic classifier G m (x) continue to expand, while the weights of correctly classified samples decrease. Therefore, the misclassified samples play a greater role in the next iteration.
  • Step 3-2-5 construct a linear combination of basic classifiers to obtain the final classifier:
  • the linear combination f (x) implements the weighted voting of M basic classifiers, the f (x) value determines the category of the instance x, and indicates the confidence level of the classification, and combines the trained weak classifier into a strong classifier to obtain the vehicle accident risk Forecasting model.
  • Step 4 Import real-time traffic data sets through VANETs to obtain the output of the prediction model.
  • C 0 means that the probability of a car accident is low or only a minor collision accident occurs
  • C 1 means that a more serious accidental injury may occur
  • C 2 indicates that the probability of a car accident is high or an accident may occur.

Landscapes

  • Business, Economics & Management (AREA)
  • Engineering & Computer Science (AREA)
  • Human Resources & Organizations (AREA)
  • Strategic Management (AREA)
  • Economics (AREA)
  • Entrepreneurship & Innovation (AREA)
  • Marketing (AREA)
  • Game Theory and Decision Science (AREA)
  • Development Economics (AREA)
  • Operations Research (AREA)
  • Quality & Reliability (AREA)
  • Tourism & Hospitality (AREA)
  • Physics & Mathematics (AREA)
  • General Business, Economics & Management (AREA)
  • General Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Educational Administration (AREA)
  • Traffic Control Systems (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)

Abstract

一种基于AdaBoost-SO的VANETs车辆事故风险预测模型,其能够为ITS和驾驶安全辅助提供理论基础。所述模型建立方法包括:首先填充研究数据集,用SMOTE算法来平衡数据集中的样本,并将每个样本特征用One-Hot编码,然后用trichotomy Adaboost-SO算法训练研究数据集获得系统模型,最后通过VANETs导入时交通数据,获得车辆事故概率。其中AdaBoost-SO是trichotomy Adaboost with SMOTE and One-Hot encoding使用SMOTE算法和一位有效编码的三分自适应提升算法;VANETs是VehicularAd Hoc Networks车载点对点网络,ITS是Intelligent Transportation System智能交通系统,SMOTE是Synthetic Minority Oversampling Technique合成少数类过采样技术。

Description

一种基于AdaBoost-SO的VANETs车辆事故风险预测模型 技术领域
本发明涉及车联网技术领域,具体涉及一种基于AdaBoost-SO的VANETs车辆事故风险预测模型。
背景技术
随着当今社会经济的发展,城市居民对旅行的便利性和舒适性提出了更高的要求,汽车数量增加,城市交通压力加大,道路安全问题也越来越严重。特别是在大城市,交通事故导致交通拥堵,并且车辆事故对人身安全的威胁越来越严重,这使得交通安全研究具有重要意义。与此同时,车载Ad Hoc网络(VANETs)作为智能交通系统(ITS)的关键技术,其快速发展具有提高道路安全性和交通效率的巨大潜力。它为有效研究道路安全提供了原始道路安全信息,并为预测车辆事故风险提供了新思路。从高度异构的资源中收集大量VANETs数据,为VANETs-Big Data的新时代铺平了道路
随着大数据和机器学习的发展,利用机器学习技术预测交通事故成为新的亮点。文献“The traffic accident hotspot prediction:Based on the logistic regression method”通过对典型因素的统计和Logistic回归分析,研究交通事故,道路类型,车辆类型,驾驶员状态,天气,日期等之间的关系,最后建立事故热点预测模型。文献“The five-factor model,conscientiousness,and driving accident involvement”和“Determining personality traits of racing game players using the open racing car simulator:toward believable virtual drivers”研究了驾驶员的 责任感与驾驶事故之间的关系,证明那些责任心强的人不太可能发生交通事故。文献“Traffic big data analysis supporting vehicular network access recommendation”开发了一种由交通大数据分析支持的智能网络推荐系统,建议车辆通过采用分析框架访问适当的网络,并使个体车辆能够基于访问推荐器自动访问网络。
然而,所有这些方法都集中在从现有交通数据中分析交通事故的原因,并且未能获得具有普遍应用价值的事故预测模型。因此,设计一个可以使用实时交通数据,并随时向车辆发出警报,为智能交通系统和驾驶安全辅助提供理论基础的车辆事故风险预测模型是很有必要的。
发明内容
本发明的主要目的在于解决现有技术中存在的问题,本发明提供了一种基于AdaBoost-SO的VANETs车辆事故风险预测模型。
一种基于AdaBoost-SO的VANETs车辆事故风险预测模型,所述模型建立的步骤包括:
步骤一:填充研究数据集;
步骤二:用SMOTE算法平衡数据集中的样本,并且将每个样本的离散特征用One-Hot编码;
具体的,将Synthetic Minority Oversampling Technique(SMOTE)算法用于解决研究数据集中每个类别的样本数不平衡的问题;
在使用SMOTE算法预处理初始研究数据集之后,可以获得每个类别的相对平衡数量的实验数据集;接下来,将每个样本的离散特征 用One-Hot编码;One-Hot编码方法为使用N比特状态寄存器来编码N个状态,每个状态具有单独的寄存器比特,并且在任何时候只有一个比特有效;
步骤三:用trichotomy Adaboost-SO算法训练研究数据集获得系统模型;
具体的,首先,构建实验数据集时,将道路安全数据随机分为训练数据和测试数据,并进行6次交叉验证,该方法充分利用了所有样本,它需要6次培训和6次测试;然后,使用trichotomy AdaBoost算法处理研究数据集;
步骤四:通过VANETs导入实时交通数据集,获得预测模型的输出;
具体的,输出值为C={C 0,C 1,C 2},表示预测对象是否属于事故发生率高;C 0表示车祸概率低或仅发生轻微碰撞事故,C 1意味着可能发生更严重的意外伤害,C 2表明车祸的概率很高或可能发生意外事故。
进一步地,所述步骤一中,具体的,在重建数据之前,找到并修改不确定或不完整的道路安全数据,以改进数据集;通常的实现方案包括填充可用特征的平均值,特殊值,类似样本的平均值,并直接忽略具有缺失值的样本。
进一步地,所述步骤二中,SMOTE算法实现过程是:
步骤2-1,对于少数类别中的每个样本x,欧几里德距离用作标准来计算与少数类别中所有其他样本的距离,以获得其k最近样本;
步骤2-2,根据样本不平衡比设置采样率N,对于每个少数类样本x,假设所选择的邻近样本是k,从其k邻近样本随机选择若干样本;
步骤2-3,对于每个选定的邻居,根据如下公式,使用原始样本构建新样本;
Figure PCTCN2019092462-appb-000001
进一步地,所述步骤三中,所述6次交叉验证的具体实施步骤如下:
步骤3-1-1,将整个研究数据集S划分为6个相同大小的相互不相交的子集;假设训练样本的数量是m,则每个子集将具有
Figure PCTCN2019092462-appb-000002
个训练样本,相应的子集为{S 1,S 2,S 3,S 4,S 5,S 6};
步骤3-1-2,将一个子集作为测试集,然后将其他五个子集作为训练集;
步骤3-1-3,通过训练数据训练模型,使用测试数据验证模型的准确性并重复六次;
步骤3-1-4,计算6个评估误差的平均值作为模型的真实分类精度。
进一步地,所述步骤三中,使用trichotomy AdaBoost算法处理研究数据集,其具体实施步骤如下:
步骤3-2-1,输入训练数据集T=(x 1,y 1),(x 2,y 2)...,(x N,y N),x i是样本的特征向量,y∈{1,2,3},本发明中使用的弱分类器是决策树;
步骤3-2-2,训练数据的权重初始化为:
Figure PCTCN2019092462-appb-000003
步骤3-2-3,对于第m次迭代,m=1,2,...,M:使用具有权重分布的训练数据集D m进行训练,获得基本分类器:
G m(x):χ→{1,2,3}
χ是要训练的数据,根据训练数据的分类结果计算G m(x)错误率,w mi表示第m次迭代中第i个样本的权重:
Figure PCTCN2019092462-appb-000004
Figure PCTCN2019092462-appb-000005
由于在每个步骤中权重被标准化,分母不需要除以样本权重的总和;
步骤3-2-4,trichotomy AdaBoost的错误率阈值e m设置为
Figure PCTCN2019092462-appb-000006
并且添加正项x,当
Figure PCTCN2019092462-appb-000007
时,保证a m≥0;根据错误率e m计算分类器G m(x)的系数:
Figure PCTCN2019092462-appb-000008
根据系数a m更新训练数据集的权重分布:
D m+1=(w m+1,1,...,w m+1,i,...w m+1,N)
Figure PCTCN2019092462-appb-000009
可以化简为:
Figure PCTCN2019092462-appb-000010
Figure PCTCN2019092462-appb-000011
其中,Z m作为归一化因子使得D m+1成为概率分布:
Figure PCTCN2019092462-appb-000012
在训练之后,基本分类器G m(x)的错误分类样本的权重不断扩大,而正确分类的样本的权重减小,因此,错误分类的样本在下一次迭代中发挥更大的作用;
步骤3-2-5,构造基本分类器的线性组合以获得最终分类器:
Figure PCTCN2019092462-appb-000013
Figure PCTCN2019092462-appb-000014
线性组合f(x)实现M个基本分类器的加权投票,f(x)值确定实例x的类别,并指示分类的置信度,将训练的弱分类器组合成强分类器以获得车辆事故风险预测模型。
与现有技术相比,本发明的有益效果是:最大迭代值为100的系统模型保证了普通道路状态下事故预测的最大精度,特殊情况下最大迭代值较小的系统模型可以提高时效性。在预测中,可以发挥系统的最大性能。
附图说明
图1为本发明所述方法的流程示意图。
图2为trichotomy Adaboost-SO模型架构。
具体实施方式
下面结合说明书附图对本发明的技术方案做进一步的详细说明。
一种基于AdaBoost-SO的VANETs车辆事故风险预测模型,所述模型建立的步骤包括:
步骤一:填充研究数据集。
具体的,在重建数据之前,找到并修改不确定或不完整的道路安全数据,以改进数据集;通常的实现方案包括填充可用特征的平均值,特殊值,类似样本的平均值,并直接忽略具有缺失值的样本。
步骤二:用SMOTE算法平衡数据集中的样本,并且将每个样本的离散特征用One-Hot编码。
具体的,将Synthetic Minority Oversampling Technique(SMOTE)算法用于解决研究数据集中每个类别的样本数不平衡的问题。所述SMOTE算法实现过程是:
步骤2-1,对于少数类别中的每个样本x,欧几里德距离用作标准来计算与少数类别中所有其他样本的距离,以获得其k最近样本。
步骤2-2,根据样本不平衡比设置采样率N。对于每个少数类样本x,假设所选择的邻近样本是k,从其k邻近样本随机选择若干样本。
步骤2-3,对于每个选定的邻居,根据如下公式,使用原始样本构建新样本。
Figure PCTCN2019092462-appb-000015
在使用SMOTE算法预处理初始研究数据集之后,可以获得每个类别的相对平衡数量的实验数据集。接下来,将每个样本的离散特征 用One-Hot编码。
One-Hot编码方法为使用N比特状态寄存器来编码N个状态,每个状态具有单独的寄存器比特,并且在任何时候只有一个比特有效。
步骤三:用trichotomy Adaboost-SO算法训练研究数据集获得系统模型。
具体的,首先,构建实验数据集时,将道路安全数据随机分为训练数据和测试数据,并进行6次交叉验证,该方法充分利用了所有样本,它需要6次培训和6次测试。所述6次交叉验证的具体实施步骤如下:
步骤3-1-1,将整个研究数据集S划分为6个相同大小的相互不相交的子集;假设训练样本的数量是m,则每个子集将具有
Figure PCTCN2019092462-appb-000016
个训练样本,相应的子集为{S 1,S 2,S 3,S 4,S 5,S 6}。
步骤3-1-2,将一个子集作为测试集,然后将其他五个子集作为训练集。
步骤3-1-3,通过训练数据训练模型,使用测试数据验证模型的准确性并重复六次。
步骤3-1-4,计算6个评估误差的平均值作为模型的真实分类精度。
然后,使用trichotomy AdaBoost算法处理研究数据集,其具体实施步骤如下:
步骤3-2-1,输入训练数据集T=(x 1,y 1),(x 2,y 2)...,(x N,y N),x i是样本的特征向量,y∈{1,2,3},本发明中使用的弱分类器是决策树。
步骤3-2-2,训练数据的权重初始化为:
Figure PCTCN2019092462-appb-000017
步骤3-2-3,对于第m次迭代,m=1,2,...,M:使用具有权重分布的训练数据集D m进行训练,获得基本分类器:
G m(x):χ→{1,2,3}
χ是要训练的数据。根据训练数据的分类结果计算G m(x)错误率,w mi表示第m次迭代中第i个样本的权重:
Figure PCTCN2019092462-appb-000018
Figure PCTCN2019092462-appb-000019
由于在每个步骤中权重被标准化,分母不需要除以样本权重的总和。
步骤3-2-4,trichotomy AdaBoost的错误率阈值e m设置为
Figure PCTCN2019092462-appb-000020
并且添加正项x,当
Figure PCTCN2019092462-appb-000021
时,保证a m≥0;根据错误率e m计算分类器G m(x)的系数:
Figure PCTCN2019092462-appb-000022
根据系数a m更新训练数据集的权重分布:
D m+1=(w m+1,1,...,w m+1,i,...w m+1,N)
Figure PCTCN2019092462-appb-000023
可以化简为:
Figure PCTCN2019092462-appb-000024
Figure PCTCN2019092462-appb-000025
其中,Z m作为归一化因子使得D m+1成为概率分布:
Figure PCTCN2019092462-appb-000026
在训练之后,基本分类器G m(x)的错误分类样本的权重不断扩大,而正确分类的样本的权重减小,因此,错误分类的样本在下一次迭代中发挥更大的作用。
步骤3-2-5,构造基本分类器的线性组合以获得最终分类器:
Figure PCTCN2019092462-appb-000027
Figure PCTCN2019092462-appb-000028
线性组合f(x)实现M个基本分类器的加权投票,f(x)值确定实例x的类别,并指示分类的置信度,将训练的弱分类器组合成强分类器以获得车辆事故风险预测模型。
步骤四:通过VANETs导入实时交通数据集,获得预测模型的输出。
具体的,输出值为C={C 0,C 1,C 2},表示预测对象是否属于事故发生率高。C 0表示车祸概率低或仅发生轻微碰撞事故,C 1意味着可能发生更严重的意外伤害,C 2表明车祸的概率很高或可能发生意外事故。
以上所述仅为本发明的较佳实施方式,本发明的保护范围并不以上述实施方式为限,但凡本领域普通技术人员根据本发明所揭示内容 所作的等效修饰或变化,皆应纳入权利要求书中记载的保护范围内。

Claims (5)

  1. 一种基于AdaBoost-SO的VANETs车辆事故风险预测模型,其特征在于:所述模型建立的步骤包括:
    步骤一:填充研究数据集;
    步骤二:用SMOTE算法平衡数据集中的样本,并且将每个样本的离散特征用One-Hot编码;
    具体的,将Synthetic Minority Oversampling Technique(SMOTE)算法用于解决研究数据集中每个类别的样本数不平衡的问题;
    在使用SMOTE算法预处理初始研究数据集之后,可以获得每个类别的相对平衡数量的实验数据集;接下来,将每个样本的离散特征用One-Hot编码;One-Hot编码方法为使用N比特状态寄存器来编码N个状态,每个状态具有单独的寄存器比特,并且在任何时候只有一个比特有效;
    步骤三:用trichotomy Adaboost-SO算法训练研究数据集获得系统模型;
    具体的,首先,构建实验数据集时,将道路安全数据随机分为训练数据和测试数据,并进行6次交叉验证,该方法充分利用了所有样本,它需要6次培训和6次测试;然后,使用trichotomy AdaBoost算法处理研究数据集;
    步骤四:通过VANETs导入实时交通数据集,获得预测模型的输出;
    具体的,输出值为C={C 0,C 1,C 2},表示预测对象是否属于事故发生率高;C 0表示车祸概率低或仅发生轻微碰撞事故,C 1意味着可能 发生更严重的意外伤害,C 2表明车祸的概率很高或可能发生意外事故。
  2. 根据权利要求1所述的基于AdaBoost-SO的VANETs车辆事故风险预测模型,其特征在于:所述步骤一中,具体的,在重建数据之前,找到并修改不确定或不完整的道路安全数据,以改进数据集;通常的实现方案包括填充可用特征的平均值,特殊值,类似样本的平均值,并直接忽略具有缺失值的样本。
  3. 根据权利要求1所述的基于AdaBoost-SO的VANETs车辆事故风险预测模型,其特征在于:所述步骤二中,SMOTE算法实现过程是:
    步骤2-1,对于少数类别中的每个样本x,欧几里德距离用作标准来计算与少数类别中所有其他样本的距离,以获得其k最近样本;
    步骤2-2,根据样本不平衡比设置采样率N,对于每个少数类样本x,假设所选择的邻近样本是k,从其k邻近样本随机选择若干样本;
    步骤2-3,对于每个选定的邻居,根据如下公式,使用原始样本构建新样本;
    Figure PCTCN2019092462-appb-100001
  4. 根据权利要求1所述的基于AdaBoost-SO的VANETs车辆事故风险预测模型,其特征在于:所述步骤三中,所述6次交叉验证的具体实施步骤如下:
    步骤3-1-1,将整个研究数据集S划分为6个相同大小的相互不 相交的子集;假设训练样本的数量是m,则每个子集将具有
    Figure PCTCN2019092462-appb-100002
    个训练样本,相应的子集为{S 1,S 2,S 3,S 4,S 5,S 6};
    步骤3-1-2,将一个子集作为测试集,然后将其他五个子集作为训练集;
    步骤3-1-3,通过训练数据训练模型,使用测试数据验证模型的准确性并重复六次;
    步骤3-1-4,计算6个评估误差的平均值作为模型的真实分类精度。
  5. 根据权利要求1所述的基于AdaBoost-SO的VANETs车辆事故风险预测模型,其特征在于:所述步骤三中,使用trichotomyAdaBoost算法处理研究数据集,其具体实施步骤如下:
    步骤3-2-1,输入训练数据集T=(x 1,y 1),(x 2,y 2)...,(x N,y N),x i是样本的特征向量,y∈{1,2,3},本发明中使用的弱分类器是决策树;
    步骤3-2-2,训练数据的权重初始化为:
    Figure PCTCN2019092462-appb-100003
    步骤3-2-3,对于第m次迭代,m=1,2,...,M:使用具有权重分布的训练数据集D m进行训练,获得基本分类器:
    G m(x):χ→{1,2,3}
    χ是要训练的数据,根据训练数据的分类结果计算G m(x)错误率,w mi表示第m次迭代中第i个样本的权重:
    Figure PCTCN2019092462-appb-100004
    Figure PCTCN2019092462-appb-100005
    由于在每个步骤中权重被标准化,分母不需要除以样本权重的总和;
    步骤3-2-4,trichotomy AdaBoost的错误率阈值e m设置为
    Figure PCTCN2019092462-appb-100006
    并且添加正项x,当
    Figure PCTCN2019092462-appb-100007
    时,保证a m≥0;根据错误率e m计算分类器G m(x)的系数:
    Figure PCTCN2019092462-appb-100008
    根据系数a m更新训练数据集的权重分布:
    D m+1=(w m+1,1,...,w m+1,i,...w m+1,N)
    Figure PCTCN2019092462-appb-100009
    可以化简为:
    Figure PCTCN2019092462-appb-100010
    Figure PCTCN2019092462-appb-100011
    其中,Z m作为归一化因子使得D m+1成为概率分布:
    Figure PCTCN2019092462-appb-100012
    在训练之后,基本分类器G m(x)的错误分类样本的权重不断扩大,而正确分类的样本的权重减小,因此,错误分类的样本在下一次迭代中发挥更大的作用;
    步骤3-2-5,构造基本分类器的线性组合以获得最终分类器:
    Figure PCTCN2019092462-appb-100013
    Figure PCTCN2019092462-appb-100014
    线性组合f(x)实现M个基本分类器的加权投票,f(x)值确定实例x的类别,并指示分类的置信度,将训练的弱分类器组合成强分类器以获得车辆事故风险预测模型。
PCT/CN2019/092462 2018-11-07 2019-06-24 一种基于AdaBoost-SO的VANETs车辆事故风险预测模型 Ceased WO2020093701A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201811319617.4 2018-11-07
CN201811319617.4A CN109558969A (zh) 2018-11-07 2018-11-07 一种基于AdaBoost-SO的VANETs车辆事故风险预测模型

Publications (1)

Publication Number Publication Date
WO2020093701A1 true WO2020093701A1 (zh) 2020-05-14

Family

ID=65865977

Family Applications (2)

Application Number Title Priority Date Filing Date
PCT/CN2019/092463 Ceased WO2020093702A1 (zh) 2018-11-07 2019-06-24 一种基于深度q网络学习的交通灯动态配时算法
PCT/CN2019/092462 Ceased WO2020093701A1 (zh) 2018-11-07 2019-06-24 一种基于AdaBoost-SO的VANETs车辆事故风险预测模型

Family Applications Before (1)

Application Number Title Priority Date Filing Date
PCT/CN2019/092463 Ceased WO2020093702A1 (zh) 2018-11-07 2019-06-24 一种基于深度q网络学习的交通灯动态配时算法

Country Status (2)

Country Link
CN (1) CN109558969A (zh)
WO (2) WO2020093702A1 (zh)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20220276063A1 (en) * 2021-03-01 2022-09-01 Mitre Corporation Method and System for Dynamically Navigating Routes According to Safety-Related Risk Profiles
CN118535944A (zh) * 2024-06-12 2024-08-23 国网湖北省电力有限公司营销服务中心(计量中心) 一种异常用电检测模型构建的方法和系统

Families Citing this family (9)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN109558969A (zh) * 2018-11-07 2019-04-02 南京邮电大学 一种基于AdaBoost-SO的VANETs车辆事故风险预测模型
CN111126868B (zh) * 2019-12-30 2023-07-04 中南大学 一种道路交通事故发生风险确定方法及系统
CN111507504A (zh) * 2020-03-18 2020-08-07 中国南方电网有限责任公司 基于数据重采样的Adaboost集成学习电网故障诊断系统及方法
CN111814836B (zh) * 2020-06-12 2022-07-19 武汉理工大学 基于类不平衡算法的车辆行驶行为检测方法及装置
CN111859291B (zh) * 2020-06-23 2022-02-25 北京百度网讯科技有限公司 交通事故识别方法、装置、设备和计算机存储介质
CN111768041A (zh) * 2020-07-02 2020-10-13 上海积成能源科技有限公司 一种基于自适应提升算法预测短期电力负荷的系统模型
CN113326971A (zh) * 2021-04-30 2021-08-31 东南大学 一种基于PCA和Adaboost的隧道交通事故持续时间预测方法
CN113780641A (zh) * 2021-08-31 2021-12-10 同济大学 一种基于迁移学习的事故预测方法及装置
CN114444765B (zh) * 2021-12-17 2025-09-16 西安理工大学 基于KNN和AdaBoost的铁路事故类型预测方法

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102984200A (zh) * 2012-09-13 2013-03-20 大连理工大学 一种适用于稀疏和稠密多种VANETs场景的方法
CN104064029A (zh) * 2014-07-07 2014-09-24 哈尔滨工业大学 一种VANETs中V2V链路延迟的动态预测方法
US20170270413A1 (en) * 2016-03-15 2017-09-21 Nec Europe Ltd. Real-time filtering of digital data sources for traffic control centers
CN108596409A (zh) * 2018-07-16 2018-09-28 江苏智通交通科技有限公司 提升交通危险人员事故风险预测精度的方法
CN108763865A (zh) * 2018-05-21 2018-11-06 成都信息工程大学 一种预测dna蛋白质结合位点的集成学习方法
CN109558969A (zh) * 2018-11-07 2019-04-02 南京邮电大学 一种基于AdaBoost-SO的VANETs车辆事故风险预测模型

Family Cites Families (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103208195A (zh) * 2013-04-08 2013-07-17 沈阳广信先锋交通高技术有限公司 多智能体交通信号控制系统
CN104112366B (zh) * 2014-07-25 2017-02-22 中国科学院自动化研究所 基于隐语义模型的交通信号优化方法
CN105677564A (zh) * 2016-01-04 2016-06-15 中国石油大学(华东) 基于改进的Adaboost软件缺陷不平衡数据分类方法
CN108154681B (zh) * 2016-12-06 2020-11-20 杭州海康威视数字技术股份有限公司 发生交通事故的风险预测方法、装置及系统
CN109544913A (zh) * 2018-11-07 2019-03-29 南京邮电大学 一种基于深度q网络学习的交通灯动态配时算法
CN109697867B (zh) * 2019-01-28 2020-10-20 深圳市欧德克科技有限公司 一种基于深度学习的交通控制方法及系统

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102984200A (zh) * 2012-09-13 2013-03-20 大连理工大学 一种适用于稀疏和稠密多种VANETs场景的方法
CN104064029A (zh) * 2014-07-07 2014-09-24 哈尔滨工业大学 一种VANETs中V2V链路延迟的动态预测方法
US20170270413A1 (en) * 2016-03-15 2017-09-21 Nec Europe Ltd. Real-time filtering of digital data sources for traffic control centers
CN108763865A (zh) * 2018-05-21 2018-11-06 成都信息工程大学 一种预测dna蛋白质结合位点的集成学习方法
CN108596409A (zh) * 2018-07-16 2018-09-28 江苏智通交通科技有限公司 提升交通危险人员事故风险预测精度的方法
CN109558969A (zh) * 2018-11-07 2019-04-02 南京邮电大学 一种基于AdaBoost-SO的VANETs车辆事故风险预测模型

Cited By (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20220276063A1 (en) * 2021-03-01 2022-09-01 Mitre Corporation Method and System for Dynamically Navigating Routes According to Safety-Related Risk Profiles
US11725955B2 (en) * 2021-03-01 2023-08-15 Mitre Corporation Method and system for dynamically navigating routes according to safety-related risk profiles
US20240003697A1 (en) * 2021-03-01 2024-01-04 Mitre Corporation Method and System for Dynamically Navigating Routes According to Safety-Related Risk Profiles
US12135218B2 (en) * 2021-03-01 2024-11-05 Mitre Corporation Method and system for dynamically navigating routes according to safety-related risk profiles
US20250060219A1 (en) * 2021-03-01 2025-02-20 Mitre Corporation Method and System for Dynamically Navigating Routes According to Safety-Related Risk Profiles
CN118535944A (zh) * 2024-06-12 2024-08-23 国网湖北省电力有限公司营销服务中心(计量中心) 一种异常用电检测模型构建的方法和系统

Also Published As

Publication number Publication date
WO2020093702A1 (zh) 2020-05-14
CN109558969A (zh) 2019-04-02

Similar Documents

Publication Publication Date Title
CN109558969A (zh) 一种基于AdaBoost-SO的VANETs车辆事故风险预测模型
CN109840660B (zh) 一种车辆特征数据处理方法及车辆风险预测模型训练方法
Ding et al. Non-linear relationships between built environment characteristics and electric-bike ownership in Zhongshan, China
CN108072381B (zh) 一种路径规划的方法及装置
CN108364467B (zh) 一种基于改进型决策树算法的路况信息预测方法
CN115100547B (zh) 基于ASPP-YOLOv4多尺度融合的无人机图像目标检测方法
Yin et al. Disentangling the effects of the built environment on car ownership: A multi-level analysis of Chinese cities
CN108470444A (zh) 一种基于遗传算法优化的城市区域交通大数据分析系统与方法
CN111598325A (zh) 基于层次聚类和分层注意力机制的交通速度预测方法
CN109887279B (zh) 一种交通拥堵预测方法及系统
CN109272745B (zh) 一种基于深度神经网络的车辆轨迹预测方法
Umer et al. Comparison analysis of tree based and ensembled regression algorithms for traffic accident severity prediction
CN118865444A (zh) 一种基于改进YOLOv8的行人小目标检测方法
CN101964061B (zh) 一种基于二类核函数支持向量机的车型识别方法
CN112035536A (zh) 一种考虑动态路网交通流的电动汽车能耗预测方法
CN114120280A (zh) 一种基于小目标特征增强的交通标志检测方法
CN110304068A (zh) 汽车行驶环境信息的采集方法、装置、设备和存储介质
CN106157624A (zh) 基于交通定位数据的多粒度道路分流可视分析方法
CN108629450A (zh) 一种液化天然气公交车尾气排放预测方法
CN118907085A (zh) 基于深度q网络的车辆风险评估方法、装置、设备及介质
Zhang et al. MaaS in bike-sharing: smart phone GPS data based layout optimization and emission reduction potential analysis
CN110119838A (zh) 一种共享单车需求预测系统、方法及装置
CN103106790B (zh) 一种基于秩和比法的平面交叉口设计变权综合评价方法
Song et al. Sign-yolo: A novel lightweight detection model for chinese traffic sign
CN116935643B (zh) 交通管理方法、装置、设备及存储介质

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 19882910

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 19882910

Country of ref document: EP

Kind code of ref document: A1