To create a prediction model for road traffic time series data(i.e vehicle count)using artificial neural network in matlab

Question

shripati 2014-2-12

0
链接

此问题的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/115881-to-create-a-prediction-model-for-road-traffic-time-series-data-i-e-vehicle-count-using-artificial-ne

评论： Tiago Goncalves 2020-3-28

I'm basically trying to develop a short term prediction model for road traffic time series data using artificial neural networks using matlab. I wish to know whether its possible to achieve the objective using matlab? And if it is possible,can someone please provide some code snippet in matlab(since i am a beginner in this field) which involves such kind of forecasting for any time series data so that i can understand how basically forecasting is done in matlab (using artificial neural networks concept)? I have the vehicle count of 5 days of a week of a particular road and wish to forecast the values of next 2 days of the week using the time series data.Hence i plan on having 5 input nodes, 5-10 neurons in hidden layer and 2 nodes in output layer as a part of my model. Any help in this regard would be highly appreciated.

1 个评论
显示 -1更早的评论隐藏 -1更早的评论

Tiago Goncalves 2020-3-28

Predictive analytics enables the ability to map-match and analyze large volumes of traffic data allows, crucial to make traffic predictions at scale. No matter how small or large the road network, the data can produce accurate traffic predictions throughout the day. It’s then possible to calculate congestion and delays based on defined free flow traffic thresholds and speed profiles with several kilometers of live trip data collected globally every day. This provides an extremely detailed predictive traffic baseline for every road, for every day of the week.

TomTom for example, achieves on route traffic prediction through a multitude range of sources around the world to generate real-time traffic information. They leverage a global community of over 600 million GPS and floating car data (FCD) probes to measure flow speeds, detect congestion and other incidents on the road in real-time.

In total, they collect traffic data from the following sources:

Smartphone partners
Popular ride-hailing and navigation applications
Automotive partners
Fleet management systems
GPS probes from TomTom connected navigation devices and smartphone apps

In this article it's possible to know more on accurate traffic prediction: What is Traffic Prediction and How Does it Work? | TomTom Blog

请先登录，再进行评论。

请先登录，再回答此问题。

Answer 1

Greg Heath 2014-2-26

2
链接

此回答的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/115881-to-create-a-prediction-model-for-road-traffic-time-series-data-i-e-vehicle-count-using-artificial-ne#answer_126279

在 MATLAB Online 中打开

You will have to include day of the week (1-7) as an input.

help narxnet

2 个评论
显示无隐藏无

shripati 2014-3-11

Suppose i have hourly traffic data(i.e count of no of vehicles passing in one hour) for 7 days of the week (between 8 am to 10 pm) and i wish to predict the traffic count for the 7 days of the next week, what should i give as the input array and the target array for my problem? Should i use NARX or NAR model for my problem? And which option would help me view the forecasted values for the next 7 days in the neural network toolbox? Help would be appreciated.

Prasad Kalane 2014-3-14