Key points are not available for this paper at this time.
Traffic-flow characteristics such as flow, density, and space mean speed (SMS) are critical to Intelligent Transportation Systems (ITS). For example, flow is a direct measure of throughput, density is an ideal indicator of traffic conditions, and SMS is the primary input to compute travel times. An attractive method to compute traffic-flow characteristics in ITS is expected to meet the following criteria: 1) It should be a one-stop solution, meaning it involves only one type of sensor that is able to determine flow, SMS, and density; 2) it should be accurate, meaning it determines these characteristics by definition rather than by estimation or by using surrogates; 3) it should preserve the fundamental relationship among flow, SMS, and density; and 4) it should be compatible with ITS, meaning it uses ITS data and supports online application. Existing methods may be good for one or some of the above criteria, but none satisfies all of them. This paper tackles the challenge by formulating a method, called the n-t method, which addresses all these criteria. Its accuracy and the fundamental relationship are guaranteed by applying a generalized definition of traffic-flow characteristics. Inputs to the method are time-stamped traffic counts which happen to be the strength of most ITS systems. Some empirical examples are provided to demonstrate the performance of the n-t method
Daiheng Ni (Fri,) studied this question.