摘要
传统的交通拥塞状态识别方法往往需要手动选择,提取特征,对于大规模和高维度的交通数据来说效率低下,难以动态地适应城市交通状态的变化,导致城市道路交通拥塞状态识别效果不佳。为此,提出一种基于机器学习的城市道路交通拥塞状态识别方法。该方法以城市道路交通视频图像作为基础数据,基于机器学习方法结合深度学习技术,自动从数据中学习特征,提高特征提取效率;通过设置交通参数与拥堵临界点后,建立城市道路交通拥塞强度评价的Logistic回归模型,通过该模型来评价当前城市道路交通视频图像内交通拥塞强度,然后将城市道路交通拥塞强度评价结果输入到机器学习算法的支持向量机模型内,再使用麻雀算法对支持向量机模型进行改进,得到最佳的支持向量机模型参数,运用该最佳参数训练支持向量机模型后,输出城市道路交通拥塞状态识别结果。实验结果表明:该方法可有效评价不同类型城市道路交通拥塞强度,并可利用机器学习算法中的支持向量机模型输出城市道路交通拥塞状态,应用效果较佳。
Traditional methods often need to manually select and extract features,which is inefficient for large-scale and high-dimensional traffic data,and it is difficult to dynamically adapt to the changes of urban traffic state,resulting in poor recognition effect of urban road traffic congestion state.Therefore,an urban road traffic congestion state recognition method based on machine learning is proposed,which takes the urban road traffic video image as the basic data.Based on machine learning method and deep learning technology,features are automatically learned from data to improve the efficiency of feature extraction.After setting the traffic parameters and the congestion critical point,the Logistic regression model for the evaluation of urban road traffic congestion intensity is established.The traffic congestion intensity in the current urban road traffic video image is evaluated by the model.Then the evaluation results of urban road traffic congestion intensity are input into the support vector machine model of machine learning algorithm,and then the support vector machine model is improved by means of the sparrow algorithm to obtain the optimal parameters of the support vector machine model.After training the support vector machine model with the optimal parameters,the identification results of urban road traffic congestion state are output.The experimental results show that the method can effectively evaluate the traffic congestion intensity of different types of urban roads,and output the urban road traffic congestion state by means of the support vector machine model in the machine learning algorithm.Its application effect is better.
作者
卞晨
BIAN Chen(Hefei University of Technology,Hefei 230009,China)
出处
《现代电子技术》
北大核心
2024年第14期142-146,共5页
Modern Electronics Technique
