摘要
城市举办大型活动期间,相关交叉口单侧交通量会急剧增加,在这种情况下准确识别交叉口的服务水平对接下来制订交叉口优化方案十分重要。本文针对该问题提出了一种小波去噪-支持向量机模式识别的方法。通过VISSIM仿真模拟交叉口单侧交通量剧增的情形获取上游的交通流数据与对应的交叉口服务水平信息,进行小波去噪与类型筛选后得到识别算法的输入数据,对比BP神经网络模型方法与支持向量机这两种常用算法的识别效果后,采纳支持向量机作为识别算法并在MATLAB上实现对交叉口服务水平的识别。与传统方法相比,该识别方法具有模型输入数据易于获取,应用时效性更强的优势。
During the major events held in the city, the unilateral traffic of related intersection will increase dramatically. How to i-dentify service level of the intersection accurately is very important to proposing the optimization program. Aiming at this problem, the paper present a recognition method of wavelet domain denoising - support vector machine. Simulated the situation of the increasing uni-lateral traffic by software VISSIM and obtained upstream traffic flow data and the corresponding service levels of intersection. After wavelet domain denoisingand data types screening , got input data for recognition algorithm. Compared the recognition efficiency of BP neural network model and Support Vector Machines, adopted SVM as the recognition algorithm, then identified the service level of inter-section in MATLAB.Compared with traditional methods, this identification method is easier to obtain input data, and it hasgreater timeli-nessin the application.
出处
《黑龙江科技信息》
2016年第16期114-117,共4页
Heilongjiang Science and Technology Information
关键词
交通量剧增
交叉口服务水平
小波去噪
支持向量机
Traffic surges
Service level of the intersection
Wavelet domain denoising
Support Vector Machines