Real-time jam prediction and intelligent management system for road traffic network area

Abstract

The invention discloses a system for implementing intelligent management based on prediction of future road traffic conditions. The system comprises a data fusion processing module, a mathematical model prediction operation module and a prediction-based traffic dispersion management module. Normalized information input is generated by using real-time acquired multi-source traffic information by a data fusion technology, the internal relation of road network traffic flow is judged through historical data, a parameterized statistical model is established, the traffic conditions (such as speed and flow) of each road section in a road network in future 90 minutes are calculated, and the time accuracy is consistent with the time interval of real-time data acquisition. Quantified traffic indexes comprising jam probability of each road section in future time, smooth traffic recovery time of the blocked road section and traffic flow abnormity alarm are further judged and generated according to the prediction result, so that decision support is provided for traffic participants and administrators.

Description

Technical field [0001] The invention is a system for intelligent management based on real-time prediction of future road traffic conditions. Background technique [0002] In the transportation domain, travel time is essential to provide navigation for travelers and dispatchers. In the traffic network, this information is usually obtained by taking the average value of each link. Based on the average travel time, we can use many shortest path algorithms to find the best travel route. A path in a traffic network is composed of one or more road segments. There are several traditional approaches to computing navigation and optimal paths for future time periods. [0003] The most common method is to take the average. This method can provide the user with an average best path. However, due to road congestion, there will be a large gap between the average travel time of a certain road segment and the real travel time at a certain time. For example, the travel time of a certai...

AI Technical Summary

Problems solved by technology

[0007] The third method is to predict by computer neural network and non-parametric regression model. This type of method has been tried by many research groups for actual traffic prediction since the 1990s. However, due to the inherent shortcomings of this method, Its calculation speed is slow and it cannot be used to predict the road network with large-scale precision coverage at the same time, and her prediction accuracy is not high. More importantly, in this type of solution, some external factors (such as weather, road zero-time construction, Large-scale activities) can not reflect the impact on traffic flow in time.

[0010] (2) Since the effects of weather or events are very different for different road segments, it is often not enough to calculate the average travel time with a single weighting factor

In addition, high-detail data for the current situation, as assumed in the first traditional method, is generally not available for most road segments, and is not effective for very short-term predictions

Images