Quality evaluation and control method of vehicle detector data on urban expressway

In order to improve the accuracy of vehicle detector data, a quality evaluation and control method of vehicle detector data on urban expressway was proposed.A data quality evaluation and control process was proposed by comprehensive considering a single location and the law of flow conservation. A data quality evaluation method based on flow cumulative curves was proposed by analyzing the characteristics of the flow cumulative curves of upstream and downstream detected sections. With the objective function that the square sum of the difference between the adjusted and the observed flow cumulative value of vehicles for all detected sections in the road network was minimum, and with the restriction condition of the characteristics of flow cumulative curves, a data quality control method based on the optimization techniques was proposed.The evaluation and control method was verified and compared by using the data collected from inductive loop detectors on a megacity urban expressway. Verified result indicates that when the introduced errors are 0.025, 0.050, 0.075, and 0.100, the mean absolute percentage errors computed by the method are 2.40%, 3.83%, 5.70% and 8.26% respectively, and its accuracy is more than 97%, so the method has better precision than comparison methods.