Traffic Congestion Carding and Optimization Modeling Analysis based on Multi-ｓｏｕｒｃｅ Data Fusion

Xinxue Jiang

Shiyuan College, Nanning Normal University, Nanning, 530226, China

Abstract: Because of the traditional mathematical model, in the traffic congestion carding ignored the instantaneous attributes of the cross section passenger flow, resulting in the carding traffic congestion degree is still high. To solve this problem, a traffic congestion carding and optimization modeling analysis based on multi-ｓｏｕｒｃｅ data fusion is proposed. Through the clear traffic congestion carding modeling foundation, the calculation of cross section passenger flow. Then the model parameters are calibrated with the cross section passenger flow as the key parameter and a cross learning mechanism of vector sharing is proposed by multi-ｓｏｕｒｃｅ data fusion. Through the representation of the vector sharing between vectors, continuous iteration, ｕｐｄａｔｅ, through the left and right parts of the alternating learning, finally get the fusion of the vector representation. The optimization modeling of traffic congestion was completed by solving the model. The results show that the designed carding model is lower than the control group in traffic congestion degree in four roads, and the designed carding model can effectively carding traffic congestion.

Keywords: Multi-ｓｏｕｒｃｅ data fusion; Traffic congestion sorting; Optimization modeling; Analysis