Research and design of ETC operation status evaluation system based on SOM-MQE

Abstract

With the continuous construction and improvement of the ETC system, the daily inspection and troubleshooting of ETC equipment have become important work to ensure the stable operation of the ETC system. In order to further improve the service level of the ETC system, this paper intends to improve the level and effect of ETC equipment inspection and maintenance work by constructing a highway ETC operation status evaluation system. By docking to obtain the physical data of ETC gantry system equipment operation and gantry business transaction data, the model based on the combination of self-organized mapping network (SOM) and minimum quantization error (MQE) is used for pattern recognition by using the quantization error of data deviation from the normal feature space. It is also applied to the whole life cycle data of the ETC gantry system to evaluate the effectiveness of equipment failure warnings. The performance of SOM, SOM-MQE, and BP are tested separately by test set data, and it is verified that the SOM-MQE model quantitatively describes the health status of the equipment by tracking the MQE value, which effectively reduces the evaluation time consumption and to a certain extent also reduces the false detection rate. In order to verify the accuracy of the algorithm in practice, this paper constructs an experimental scenario with different traffic flow as the environmental parameter to verify the practicality of the system in the actual environment, relying on the actual high-speed project. The overall state evaluation accuracy rate is close to 94%, which has the ability to improve the comprehensive protection of ETC gantries.