Through a characteristic analysis of multi-sensors intelligent transport system（ITS） data fusion on data layer and neural network

a multi-sensors ITS data fusion approach using the genetic algorithm and the improved BP networks algorithm on the base of their advantage and disadvantage is proposed

and the implementation processes is designed.The data fusion approach applied to eight-flow data from crossway demonstrates that the proposed fusion approach can not only save the training time greatly

but also process the data quality control effectively and improve the level of the data accuracy.