Road Network Reliability Evaluation Based On Floating Car Data

Road network reliability is fast developed as a new technology to evaluate the transportation network. Many scholars try to find new methods to fit the reality of road network. GIS (Geographic Information System) is being more widely used in digital city and other related research area with its powerful spatial analysis function, combination of GIS and traffic analysis is one direction of transportation planning development.Urban resident travel behavior can reflect urban traffic demand characteristics, travel behavior regularity and traffic flow distributing; therefore, it can provide decision-making for transportation planning. The conventional traffic survey has shortcomings of high cost, small sample, poor quality, and untimely; it cannot dynamically model urban resident's travel behavior characteristics. Actually, resident's travel information and road traffic condition are dynamic, but there are few studies on urban resident travel behavior characteristics dynamic modeling and its experimental results.As a main mode of non-daily traveling, taxi plays an important role in urban public traffic system. Its driving characteristics can partly reflect the travel behavior of urban resident. Nowadays, many taxi companies have established GPS monitoring and dispatching system. Therefore, we can analyze urban resident travel behavior characteristics based on FCD (floating car data) of taxi GPS.On the basis of the above, urban resident travel behavior characteristics modeling which based on GIS traffic analysis method and employed FCD of taxi GPS in Shenzhen city as survey sample. Traditional network reliability evaluation methods are often based on the static structure or mathematical theory; they can not reflect actual operation of road network because of lack of accurate data to support. In our research, urban resident travel behavior characteristics based on FCD are modeled, and thus they are used for evaluating road network reliability, it is a greatest innovation of this paper. Linear coefficient is a main index of measuring whether roads are convenient or not, traditional linear coefficient only reflects the geometric properties of road networks, traffic flow is not considered. But in reality, traffic flow has a deep impact on road network reliability. Therefore, this paper presents DNC (Dynamic linear coefficient) which loads traffic flow to calculate road network reliability; it is an improvement to evaluate road network reliability relative to traditional method.