A Combination Algorithm of Freeway Traffic Automatic Incident Detection

Traditional methods to detect freeway accidents are simple in theory and practical in operation, but would usually fail to deliver high detection rates and low rates of false alarm. Scholars and researchers, both at home and abroad, are shifting their attention to complicated algorithms (such as Neural Network algorithm). Though improved effectively in theory, these algorithms are less practical, due to their massive data transfer, complicated data processing, and strict requirements for equipment. Therefore, how to build a practical algorithm that is suitable for the current freeway detection equipment and software systems, with the ability to achieve accurate detection results, has triggered a heated discussion in the field of freeway accident detection technologies. Based on the objective of lowering hardware cost and operation expenses and improving detection results, this paper offers an algorithm that combines California algorithm and filter algorithm, by grasping the change patterns of eigenvalues in each algorithm. From simulation studies it is shown that, compared with the single application of either California algorithm or filter algorithm, the combination algorithm can enhance detection rates and effectively reduce false alarm rates without increasing software or hardware expenses.